Selective Deprivatization among American Religious Traditions: The Reversal of the Great Reversal*

Selective Deprivatization among American Religious Traditions: The Reversal of the Great Reversal* MARK D. REGNERUS, University of North Carolina at Chapel Hill CHRISTIAN SMITH, University of North Carolina at Chapel Hill Abstract From the advent of this country, religious traditions and movements have displayed remarkable patterns of both withdrawal from and engagement with American public life. We highlight this history, present a theory about selective deprivatization in the past and present, and test for the current presence of religious privatization and deprivatization in the lives ofamericans. Our results conclude that a significant minority ofamericans resist individual-level privatization. They want religion to speak to social and political issues, and act accordingly. Among religious traditions, conservative movements such as evangelical Protestantism are the most publicly oriented, constituting a reversal of pastgenerations. Liberal Protestants, once the most powerful religious voice in public arenas, are now much more privatized than conservative traditions. Religion's place in American public life is a matter greatly contested in public discourse today. A sociological approach to the history of Christianity in America suggests that a key to the phenomenal strength and growth of evangelical Protestantism in the early national period was its separation from local and federal establishment (Finke & Stark 1992). They sought souls, not political power, and because of this they overwhelmingly eclipsed mainline denominations in size. Is *We gratefully acknowledge the support of a grant from the Pew Charitable Trusts. An earlier version of this article was presented at the annual meeting of the Association for the Sociology of Religion in New York, August 1996. We are indebted to Judith Blau, Craig Calhoun, David Sikkink, Brian Steensland, Ray Swisher, Robert Woodberry, and two anonymous reviewers for helpful comments on earlier versions. Direct correspondence to Mark Regnerus, Department of Sociology, University of North Carolina, CB #3210, Chapel Hill, NC 27599-3210. E-mail: Mark_Regnerus@unc.edu. The University of North Carolina Press Social Forces, June 1998, 76(4):1347-72

1348 / Social Forces 76:4, June 1998 the same approach true of evangelical Protestantism today? And what of liberal Protestants and Roman Catholics, who as recently as the 1980s amassed over one hundred thousand supporters to protest Ronald Reagan's Central American policies (Smith 1996b)? No doubt Americans display high levels of religiosity when compared with their European cousins, from whom much of their religion came. But what does personal religiosity mean for American public life? We address in this article the relationship of religion to public life in America today, and will answer the following four questions: 1. What do religious privatization and deprivatization involve? 2. What predicts the likelihood of religious privatization and deprivatization? 3. Are intensively religious traditions inversely related to public (or deprivatized) religion, as they once were? 4. How are American religious traditions deprivatizing today? We answer these questions by looking at American religious history through lenses of a theory of selective religious deprivatization, and then examining what Americans currently believe about the proper role for religion: should it be a private matter, kept out of public debates about social and political issues, or a public matter, engaged in such debates? Following this, we look at types of public activities in which religious Americans are currently involved, in order to assess if attitudes favoring deprivatization translate into public actions. Public Religion in American History Religion's role in the public/private distinction has always been ambiguous (Casanova 1994:40; Williams & Demarath 1991:418; Wuthnow 1989b:107). In one sense, religion is unarguably located in the private sphere, since the freedom of its expression is a basic individual right, protected from state interference. Religious organizations receive nearly all their support from private sources (Wuthnow 1989b: 110). On the other hand, American religious history is a patchwork quilt of religious movements seeking public influence, political movements appropriating religious resources and symbols, and denominations and sects shifting back and forth alternately seeking a public voice and condemning the same (Casanova 1994:41). Prior to the legal separation of church from state provided for in the Bill of Rights, variants of Protestant religion were intertwined with colonial government and culture. Commenting on Puritan settlers, Wuthnow (1987: 86) captures the close relationship: "The new land would remain free, (they) often reminded themselves, only as long as the religious ideals on which their settlements were founded were strictly and collectively obeyed;' typically "on pain of expulsion or threat of divine punishment." Seeking to unite a geographically and theologically diverse America, however, the First Amendment provided a framework by which

Selective Deprivatization and American Religious Traditions / 1349 citizens could be assured that the new federal government would not support any one religion above others. While-officially disestablished, varieties of Protestantism nevertheless enjoyed a religious and cultural hegemony for well over 100 years following the adoption of the Bill of Rights. State establishments of Protestantism lingered as late as 1833 (Fowler & Hertzke 1995). It was not until nearly the Civil War, however, that the public/private dichotomy came to be clearly distinguished from a question of how congregations were financially supported, and attached instead to dissenting voices and actions in the public sphere. At that time, figures emerged within evangelical and mainline Protestantism to connect religious faith to the abolition of slavery. Individuals as diverse as Lyman Beecher, his daughter Harriet, William Garrison, and Charles Finney gave the abolitionist movement a decidedly religious tone. Southern evangelical and mainline Protestantism, however, preferred stricter separation of church and state, protecting their religion from questions of political and social import. As it became clear that capitalist industrialization and urbanization were making Americans less, not more, moral (Handy 1984; Casanova 1994), public religious voices emerged near the turn of the century to protest its human abuses. Those voices largely belonged to urban, liberal Protestant elites, whose postmillennial theology moved them to confront factory owners, tenement landlords, and city politicians in what has come to be called the Social Gospel movement. To the extent that they would "hold fast to moral and reformatory work, Roman Catholic partnership was welcomed (White & Hopkins 1976: 226). Evangelicals and their fundamentalist cousins were largely absent from this movement, save for the Salvation Army and the rescue (as distinct from reform) mission movement. Religious historians identify this period of the early twentieth century as the Great Reversal: the U-turn of northern evangelicalism from earlier public (and progressive) positions on such issues as slavery, women's suffrage, prison reform, and child labor, to a policy of privatized faith that concentrated on a personal evangelism rooted in the premillennial conviction that the world was inevitably worsening (Marsden 1980; Moberg 1977). Southern evangelicals and fundamentalists, briefly venturing forth from the private sphere to wage the infamous creation-evolution struggle, promptly withdrew again. Liberal Protestantism again made its public presence felt during the Vietnam conflict and the civil-rights movement. The black church, largely marginalized from public power up to that point, spearheaded the latter. Evangelical and fundamentalist Protestantism, whether because of their own strict policy of separation or their satisfaction with the scope and character of American culture and policy, or both, were largely silent. But this all changed in the late 1970s. In light of the rise and fall of the Moral Majority, an evangelical president (Carter), a friend-of-evangelicals president (Reagan), the rise of the Christian Coalition, the explosion in conservative religious lobbies, the politicization of evangelical

1350 / Social Forces 76:4, June 1998 ministers and figureheads such as Pat Robertson, James Dobson, and Ralph Reed, and the new salience given such issues as abortion, sexual orientation, race policies, taxation, and free-market economics, it is evident that conservative Protestants shifted away from the private side of the spectrum (in the form of the New Christian Right) at least as measured at the ballot box (Guth & Green 1991). The language of political responsibility appeared to replace that of premillennial cynicism (Lienisch 1993). Less evident at this point, however, is whether this movement includes social in addition to political engagement. Rethinking the Privatization Debate Sociological theories of religion have had a difficult time coming to grips with the American phenomenon of shifting private/public commitments among religious groups. Part of this problem lies with sociologists' tendency to overlook the multiple levels upon which secularization and privatization can occur: institutional differentiation, organizational evolution, and individual-level decline in beliefs and practices (Chaves 1989, 1994; Yamane 1997). For years, privatization was understood as one aspect of a longer, longitudinal process of secularization (Berger 1967; Luckmann 1967) that held sway among sociologists of religion. It was conceived as a societal, institutional force an irresistible monolith whose effects would be felt at all levels. Indeed, the evidence for societal-level privatization is strong, though understudied (for examples, see Wuthnow 1988; Marsden 1994). Nevertheless, religion is far from privatized in all respects. Indeed, religion is deprivatizing on many fronts (Casanova 1994). Public religion is not an oxymoron, nor has ever been one. Weber (1986: 528, 578-79) argued that the content of otherworldly religion, while providing escapist (private) freedom for individuals, may also demand public "liberation from the senseless treadmill and transitoriness of life;' prompting tensions with the social and political institutions of the world. The traditional theory of complete privatization is a problematic one, then, and needs to be dropped. In its place, sociologists must specify their analyses of privatization and deprivatization (to use Casanova's term) by clarifying the levels at which privatization and deprivatization occur, examining historical patterns and crosssectional portraits, and actual and residual effects. To clarify our own intentions in this article, we analyze deprivatization at the individual and organizational levels.' A Theory of Selective Deprivatization Wuthnow (1983: 174), in reviewing the political rebirth of American evangelicalism, asks the essential question: "How is it that the same religious convictions can so thoroughly discourage political activity at one moment and only

Selective Deprivatization and American Religious Traditions / 1351 a short while later promote it so enthusiastically?" The same question can and should be asked about all American religious traditions.2 Part of the answer, we suggest here, lies in understanding the triangular relationship between powerful sociohistorical events or periods, specific religious traditions (cultures), and their adherents. Borrowing from Swidler's (1986: 278) exploration of how culture influences action, "individuals in certain phases of their lives, and groups or entire societies in certain historical periods, are involved in constructing new strategies of action." In such "unsettled" periods, ideologies and here she includes religious meaning systems "establish new styles or strategies of action;' (Swidler 1986:278). Or, to better clarify the agency Swidler apparently intends, it would be more appropriate to say that religious producers (leadership and laity, not the ideologies themselves) are periodically compelled by social upheavals to refashion aspects of their own religious cultures. Under stress, they create new means to old, or new, cultural ends. In somewhat similar terms, Wuthnow (1987) himself points toward an answer to his question above. Disturbing "shifts in the moral order;' he suggests, "result in the appearance of ideological movements;' including innovative ideologies from older, traditional sources such as religion (Wuthnow 1987:152, 1989a). Though Wuthnow exhibits a more structuralist approach than Swidler, both arrive at similar conclusions concerning the development of ideological (including religious) change. What is more, religious traditions can often institutionalize ideological changes and make them seem "taken-for-granted" more successfully than newer movements, since they enjoy both the weight of hundreds to thousands of years of historical continuity and precedence as well as the luxury of remarkably "polysemous resources" symbols, language, and texts which can hold many and contradictory meanings (Sewell 1992). To leap from these theories to history, American Protestantism (and arguably most Americans) experienced the immediate pre- and post-civil War upheaval, the Industrial Revolution, the post-great War disillusionment, the Great Depression, and the cultural revolution of the 1960s, among other periods, as "unsettling" eras, or disturbing shifts in the moral order. Such times prompted a revisiting of traditional or commonsense strategies of action by the organizational sources of political and religious meaning systems, if for no other reason than to redress disgruntled electorates and empty pews. 3 It is during these "unsettled" times that religious traditions (such as evangelicalism or liberal Protestantism) look more like active ideologies than settled, taken-for-granted ways of life (Swidler 1986). During these cultural upheavals, ritual practices, doctrines, and ideas about engagement with the broader culture are likely to change form. In addition, religious traditions seldom change these aspects in step with fellow religious traditions ever their competition. Religious responses to the Industrial Revolution provide dynamic examples: liberal Protestant and institutional Catholic doctrine began displaying more humanist concerns, labor politics, modernist Christology and soteriology, and calls for religious engagement with decadent,

1352 / Social Forces 76:4, June 1998 exploitative "society." Conservative Protestants reacted with diametrically opposed positions, bearing evidence of the selective nature of deprivatization. A similar pattern followed the Great Depression, years after the Social Gospel had had its heyday (Handy 1960). It is theoretically and historically plausible, and intuitively probable, then, that the cultural revolution4 of the 1960s has prompted the most recent "unsettled" period in American life, leading religious traditions to rethink their assumptions and practices. Such nascent ideologies, to borrow Warner's (1988) term, are reactionary in nature always lagging several years behind the "unsettling" events. Such a lag may have occurred when fundamentalist and evangelical Protestantism did not politically emerge during, or immediately after, the Vietnam war. Their contemporary active politicization must be understood in light of the plausible idea that America (or at least many Americans) still inhabits the "unsettled" era, or the immediate wake thereof, prompted by the cultural upheaval surrounding the 1960s. If conservative Protestants are in fact currently mobilizing politically and socially for religious reasons, then the model outlined above predicts an alternate privatization among more liberal Protestant traditions. We will shortly assess whether this is the case. In sum, privatization and deprivatization are lagged, and varying, responses by religious traditions to periodic unsettling eras. While less helpful for explaining historical patterns of privatization and deprivatization, Casanova (1994: 57-58) provides important clues for its existence by distinguishing three purposes for which religion might deprivatize: 1. To protect not only its own freedom but all modern freedoms and rights. 2. To question and contest the absolute lawful autonomy of the secular spheres without regard to extraneous moral and ethical consideration. 3. To protect the traditional lifeworld from administrative or juridical state penetration. Lovin (1986) offers a similar, though more general, schema. His three reasons of guaranteeing freedom, providing social justice, and maintaining order correlate with Casanova's trichotomy. With these in mind, it is historically evident that not only is deprivatization a selective phenomenon among religious groups, but that religious groups have distinctly different purposes and goals for their periodic deprivatization. There is less variation with respect to means: religious traditions seldom use force or employ revolutionary tactics. Following Casanovas lead, then, liberal Protestants and progressive Catholics typically deprivatize for the second reason listed above. Their participation in the Social Gospel movement, the civilrights movement, and the Central America peace movement all exhibit these characteristics, in keeping with their intent to extend "justice, equality, and human values" to those who lack them (Warner 1988: 293). Conservative Protestants, on the other hand, deprivatize for Casanova's third reason: resisting state penetration into the lifeworld. From resisting evolutionism to fighting blue laws, from abortion and sexual orientation legislation to welfare

Selective Deprivatization and American Religious Traditions / 1353 reform, divorce law, and tax breaks, conservative Protestants and Catholics past and present wish to protect their own from lifeworld invasions that appear to undermine moral status quos. The nuclear family in particular has become the contemporary symbolic battleground for resistance (Luker 1984). We will revisit Casanova's thesis in the discussion section. While we have applied Swidler's (1986), Wuthnow's (1988; 1989a), and Casanovas (1994) theories on ideology, cultural change, and religious action to American religious history, the literature is already dense and contradictory on current social and political involvement especially on the part of religious conservatives. Some suggest it is evangelicals who are deprivatizing as they become more educated, move from small towns to cities and suburbs (primarily in the South), yet struggle to maintain the religious and moral values they "learned at the creek bank" (Balmer 1996: 77; see also Lienisch 1993 and Schmalzbauer 1993). Guth (1981) argues that the forces of modernization and secularization are responsible for not only moving evangelicals into the middle class but also raising their public consciousness and ire and equipping them with the cultural resources to mobilize. Students of the New Christian Right, however, postulate that their public support is greatest among less-educated working-class evangelicals and fundamentalists (Wilcox 1989). Even breaking down Protestant groups according to conservative denomination, literal approach to the Bible, and charismatic identity has proven to be of positive predictive value for some forms of political action and support (Kellstedt & Green 1993; Kellstedt & Smidt 1993; Wilcox, Jelen & Leege 1993). On other forms of action, such as protesting and demonstrating, religious affiliation and biblical literalism appears to have a negative effect (Sherkat & Blocker 1997). However, closer examinations of specific social movements can yield a mixed or positive interaction with religious affiliation (McAdam 1988 and Smith 1996a, respectively). Finally, still others posit that religious beliefs and organizations themselves have less agency in a religion's deprivatization and mobilization than the declining economic and social status of its adherents. Status politics, it is argued, infused conservative Protestants' push for Prohibition (Gusfield 1963) as well as the eradication of evolutionism in public schools. Today status politicking is argued to be behind the emergence of the New Christian Right (Lipset & Raab 1981), as well as fundamentalism everywhere (Thurow 1996). In studying the demographics of the privatized, Greer and Roof (1992) found youth, residents of the Northeast and West, as well as whites and males, to be the most privately religious. Nevertheless, few of these studies take a comprehensive approach to public engagement, exploring both political and social involvement. Most concern themselves solely with the political rebirth of religious conservatives. And none explores religious reasons for engagement. Indeed, most of the research cited above has focused on latent conditions in existing social structures interpretive snapshots neglecting the role of history and ideology.

1354 / Social Forces 76:4, June 1998 Analytic Approach In keeping with our intent to answer the four questions stated at the outset, we have chosen to take a nested model approach to the key dependent variable about the public or private role for religion. Specifically, model 1 includes only religious identities, and the five successive models add demographics, religious controls for charismatic, denominational affiliation, biblical literalism, and six interaction effects. We do this in order to establish exactly which religious traditions (as measured by self-identity) are currently deprivatized, which religious and demographic factors affect this, and in what order. We have also added interaction terms where theoretical arguments discussed above have merited their inclusion. We took a similar, though truncated, approach to the five "public action" dependent variables and also included the original dependent variable (private or public role for religion) as an independent variable. The results, therefore, paint the most explicit portrait possible of which religious traditions legitimate deprivatization and which traditions are presently most publicly active. We used logistic and ordered logit regression to estimate the models. Data We analyze data from the 1996 Religious Identity and Influence Survey, funded by the Pew Charitable Trusts. This is a cross-sectional, nationally representative telephone survey probing the religious beliefs, identities, and behaviors of Americans. In addition to standard measures of religiosity and denomination, the survey probed respondents' ideas about education, gender issues, morality, and involvement in society. The telephone survey was conducted by FGI, Inc., a national survey research firm based in Chapel Hill, North Carolina, from January to March 1996. Survey Sampling, Inc., provided the randomly generated sample, which is designed to sample all telephones in the U.S., excluding Alaska and Hawaii. The research design included ten calls for each number, and three callbacks to convert refusals. In order to randomize responses within households, and so to ensure representativeness by age and gender, interviewers asked to speak with the person in the household who has the next birthday. The sample population was Americans over the age of seventeen years, with an oversample of Protestants. And of those who were identified as Protestants, only those who said they attend church at least 2 times a month or who said their faith was "extremely important" in their lives were interviewed. For comparison purposes a smaller set of respondents who did not pass the screen were also interviewed. A total of 2,087 interviews with churchgoing Protestants was completed, as well as 504 interviews with respondents who did not pass the screen. Since churchgoing

Selective Deprivatization and American Religious Traditions / 1355 Protestants were oversampled, weighting procedures have been applied to all descriptive statistics and regressions. The response rate for all respondents was 69%. Dependent and Independent Variables Our key dependent variable measures people's attitude about the role of religion: does the respondent believe that religion should be a private matter, one that should be kept out of public debates over social and political issues, or not? It is dichotomous (0 = private matter, 1 = public). We also model religious and demographic effects on five public actions. Each of the actions except voting is ordered, from one to three according to the level at which respondents indicated their participation (1 = none, 2 = some, 3 = a lot). The actions include written, called, or visited an elected official (lobby); participated in public protests or demonstrations; volunteered for a church program that serves the local community; and volunteered for a local community organization not related to the church. Voting is a dichotomous measure (0 = do not usually vote). Among the independent variables, we distinguish ten mutually exclusive religious identities. Each is coded dichotomously. The first five are self-identified categories of churchgoing Protestants: fundamentalist, evangelical, mainline Protestant, theologically liberal Protestant, and other Protestant (a category for those respondents who either did not know or did not identify with the first four categories). Respondents were given the chance to identify with each of these labels, and if more than one was selected, had to choose the identity which best described themselves. If multiple identities were chosen, we coded only this "best" descriptor or identity. The reliability of these five self-identities is enhanced by high levels of religiosity. As mentioned above, these respondents were required to have either a fairly high degree of church attendance (at minimum twice per month), or have their faith be "extremely important" to them. Nominally religious respondents are not among these five self-identities. They are coded as such according to the above two religiosity measures (low in attendance and their faith is not that important to them), and grouped in self-identified categories of Protestant and Catholic. Practicing Catholics are simply self-identified Catholics who attend church at least twice per month. Non-Christian religious are a small category including Jews, Muslims, and Buddhists, among others. Finally, the nonreligious category includes those who do not identify with any religious tradition. We believe that this classification system is an improvement over previous strategies of religious measurement because it reflects how people think about themselves, respects the interdenominational reality of religious movements, and allows for finer distinctions. Nevertheless, we recognize its novelty and have given

1356 / Social Forces 76:4, June 1998 in Appendix A a table cross-classifying the religious self-identities by their denominational affiliation and biblical literalism. As we mentioned above, we are interested in other religious factors that might affect the dependent variable, and the order of this effect. A causal order of various aspects of religion on its deprivatization has neither been established nor studied. 5 Each of the variables is dichotomous. The first measures whether the respondent identifies himself or herself as a charismatic Christian or with the charismatic movement. The second religious control variable is denomination, for which we use Smith's (1990) GSS system for classifying Protestant denominations. The variable is a dichotomous measure of a respondent's affiliation with a "conservative (or fundamentalist) Protestant" denomination and a "liberal Protestant" denomination, a control used extensively in other studies. The third control is theological: how the respondent interprets the Bible. Biblical literalism is often used as a measure of conservative Protestantism (Dixon, Jones & Lowery 1992). Nevertheless, a literal approach is no longer a mark of evangelicalism as it is for fundamentalism (Ammerman 1982; Hunter 1983,1987, Marsden 1980). This variable has four ordered categories, the first three of which all presume (via an answer to the previous question) the Bible to be the inspired word of God: (1) The Bible is true in all ways, and to be read literally, word for word. (2) The Bible is true in all ways, but not always to be read literally. (3) The Bible is true primarily about religious matters, but may contain errors about other things. (4) The Bible is not the inspired word of God. We code the first answer as a literalist approach. Finally, we include several demographic controls. The following variables are dichotomous in the models: race (one each for both "black" and "other nonwhite"); sex (1 = female); education (1 = college graduate); region of residence (1 = South); and financial decline (1 = present financial situation is worse than it was ten years ago). The rest are continuous or ordinal controls: age (actual years), county size (census categories on 13-point scale), and income (an 11-point scale comprised of $10,000 increments up to $100,000 beyond this is an open category). Results Table 1 displays percentages of support for religion's public voice, one that should speak to social and political issues. There is a striking difference between traditionally conservative Protestants and their more liberal cousins here. While 68% of evangelicals and 63% of fundamentalists think religion should speak to public issues, only 48% of mainliners and 37% of liberals do. Among practicing and nominal Catholics, the numbers dip even lower, as they do with nominal Protestants and non-christians. Between the races, 34% of whites believe religion ought to be a public matter, contrasted with 48% of blacks. Those with a college

Selective Deprivatization and American Religious Traditions / 1357 TABLE 1: Frequencies of Support for Religion as a Public Matter, One That Should Speak to Social and Political Issues Religious Identity Percent of N Percent Saying "Public" Total College Educated Characteristic Percent Saying "Public" Churchgoing Protestant Fundamentalist 6.4 62.5 62.4 Male 32.2 Evangelical 6.9 68.3 78.1 Female 36.2 Mainline 9.0 48.3 44.3 Liberal 6.7 37.3 23.1 Black 48.4 Other 4.0 55.4 56.4 White 33.7 Nominal Protestant 21.5 26.5 20.0 Roman Catholic College education 29.7 Nominal 15.5 11.8 11.5 No college education 36.6 Practicing 8.8 34.5 25.6 Non-Christianreligious 8.1 23.3 16.0 Southerner 39.7 Nonreligious 13.2 15.5 6.9 Non-Southerner 33.1 Allpersons 34.4 29.7 degree were 7 percentage points less likely to support public religion than those without the degree. However, the intersection between the religious identities and education is substantial: while only 23.1 % of educated liberals think religion to be public, fully 78.1% of educated evangelicals do. When comparing overall percentages with college-educated percentages, only among evangelicals was there an increase in publicness (9.8%) of greater than 1%. The privatizing influence of education is evident among nearly every other religious identity. Table 2 shows estimated odds ratios of supporting religion as a public matter instead of a private matter, one that should speak out in debates about social and political issues. All six models are significant, though the additional variables from Models 2 through 4 do not add statistical significance in keeping with the additional degrees of freedom they add, suggesting primarily the marginal explanatory contribution of adding the demographic variables to the model. The real story then, appears to be in the religious self-identities. In all six models, four religious selfidentities - mainline Protestants, "other" Protestants, fundamentalists, and evangelicals - are significantly more supportive of religion as a public matter than theologically liberal Protestants. This is most clearly the case among conservatives - the odds of evangelicals supporting a religious faith that speaks publicly are 3.7 times that of liberal Protestants in Model 1. Fundamentalists follow them at an odds ratio 2.9 times that of liberal Protestants. In reverse, the odds of

1358 / Social Forces 76:4, June 1998 TABLE 2: Estimated Odds Ratios of Supporting Religion as a Public Matter, One That Should Speak to Social and Political Issues$ Independent Variable Modell Model2 Model3 Mode14 Model 5 Model 6 Religious Identitiesb Fundamentalist 2.901*** 2.982* 2.904*** 2.735*** 2.582*** 2.149** Evangelical 3.713*** 4.845*** 4 494*** 4.192*** 4.042*** 2.841 Mainline Protestant 1.600* 1.899** 1.952** 2.043** 2.073** 1.828* Other Protestant 2.099*** 2.377** 2.434** 2.373** 2.299** 2.321** Nominal Protestant 594**.652*.700.568*.657.647 Practicing Catholic.899 1.118 1.201.968 1.026 1.013 Nominal Catholic.234***.286***.308***.250***.275***.268*** Non-Christian but Religious.855 1.125 1.217 1.039 1.219 1.275 Nonreligious.314***.320*** 345***.280***.328***.322*** Demographic Controls Black 1.440 1.418 1.375 1.344 1.344 Other Race/Ethnic Group.776.769.761.716.729 Female.774*.769*.770*.775*.785* Ten-year Financial Decline.964.957.947.958.942 Education.701**.700**.714**.746*.595*** Age.991** 991**.991**.991**.991* Income.962.963.963.968.968 Resident of the South.983.983.981.968.941 County Size.953.952*.952*.954.953 Religious Controls Charismatic 1.774* 1.707* 1.659* 1.647* Conservative Protestant Denomination.872.835.823 Liberal Protestant Denomination.559*.581*.567* Biblical Literalism 1.525** 1.548** Interaction Effects Fundamentalist Education 1.848 Evangelical * Education 3.928**

Selective Deprivatization and American Religious Traditions / 1359 TABLE 2: Estimated Odds Ratios of Supporting Religion as a Public Matter, One That Should Speak to Social and Political Issuesa (Continued) Independent Variable Modell Modell Model3 Mode14 Model5 Model6 Mainline * Education 1.544 Fundamentalist* Financial Decline 1.848 Evangelical * Resident of the South 1.289 Evangelical * County Population.992 Intercept.585*** 2.970* 2.783* 3.377* 2.747 2.940 N 2,590 2,174 2,174 2,174 2,174 2,174 Pseudo Rz.132.160.169.178.187.197-2LogL 319.96*** 300.91*** 307.02*** 313.30*** 322.27*** 334.91*** a From Logistic regression procedure. b Identity odds ratios are compared to theologically liberal Protestants. C Denominational odds ratios are compared to the moderate Protestant denominations *p<.05 **p<.01 ***p<.001 nominal Protestants, nominal Catholics, and the nonreligious supporting public religion hover well below that of liberal Protestants. All other identities were insignificantly different. Models 3 through 5 reflect the addition of religious control variables to test their effect and causal order with the religious identities. Identifying oneself as a charismatic or with the charismatic movement significantly predicts support for religion as a public matter. This remained the case, strongly, through all successive models. Membership in a conservative Protestant denomination proved insignificantly different from membership in a moderate Protestant denomination. A liberal Protestant denomination, however, predicted significant privatization when compared with moderates. Like charismatic identity, biblical literalism is also a positive predictor as compared with other biblical interpretive approaches. The inclusion of the religious controls diminishes the evangelical and fundamentalist self-identities' odds ratios, though by marginal amounts. Even when holding constant these four religious controls, identifying oneself as an evangelical elicits an odds ratio four times that of liberal Protestants on the dependent variable. With respect to the demographic variables, only sex, education, age, and sporadically county size appear as significant demographic factors each in a privatizing direction. Education appears to have the strongest effect. None of these

1360 / Social Forces 76:4, June 1998 TABLE 3: Estimated Odds Ratios of Participation in Public Actionsa Independent Variable Protest/Demonstrate Model 1 Model 2 Lobby Elected Official Modell Model2 Voting Modell Modell Religious Identitiesb Fundamentalist.713.610 1.453 1.317 1.029.849 Evangelical 1.054 1.096 1.691* 1.457 2.155* 1.439 Mainline Protestant.424*.525 1.336.929 2.156*.990 Other Protestant.420.492.675.685.734.559 Nominal Protestant 544*.463* 1.078 1.009.828.580 Practicing Catholic.743.782 1.194 1.193 1.751* 1.175 Nominal Catholic.843.807 1.104 1.063 1.461 1.509 Non-Christian but religious.521.511 1.096 1.013.370***.234*** Nonreligious 1.135.713.775.704.341***.323*** Religious Control Religion as public matter 1.386* 1.059 1.130 Interaction Effect Evangelical*education 1.048.877.751 N 2,582 2,169 2,578 2,169 2,590 2,174 Pseudo R2.385.441.296.391.078.261-2LogL 24.07** 116.88*** 29.54*** 224.72*** 145.32*** 323.76*** variables was significant when demographic effects were regressed alone (model not shown). Each became so with the addition of religious self-identities. Neither income nor self-assessed decline in income, nor residency in the South, significantly affected the dependent variable. Model 6 includes six interaction effects, the inclusion of which was based on theoretical assertions discussed earlier in the article. Of these six, only one proved significant - that between the evangelical self-identity and education (college degree). The magnitude of the odds ratio (3.9) and the resultant insignificance of the original evangelical variable suggests a very important interaction here: educated evangelicals are the ones among all evangelicals who most want religion to speak to public issues. Education, which significantly predicts privatization in models 2 through 6, has the reverse effect when interacted with the evangelical self-identity. Table 3 displays truncated versions of the nested regression models in Table 2. Models of the five public actions - protesting, lobbying, voting, and religious and nonreligious community voluntarism - indicate the relative levels of participation on the part of the religious self-identities, and how, if at all, the support for public religion is translated into acts in the public sphere. Participation in each action is

Selective Deprivatization and American Religious Traditions / 1361 TABLE 3: Estimated Odds Ratios of Participation in Public Actionsa (Continued) Voluntarism (Religious) Voluntarism (Nonreligious) Independent Variable Modell Model2 Modell Model 2 Religious Identitiesb Fundamentalist Evangelical Mainline Protestant Other Protestant Nominal Protestant Practicing Catholic Nominal Catholic Non-Christian but religious Nonreligious Religious Control Religion as public matter Interaction Effect Evangelical*education N Pseudo R2-2 Log L 1.300 1.200.740.863 1.803** 1.523.909.887 1.193 1.069.953.891.751.726.574*.634.263***.267***.706*.765 1.025 1.022 1.250 1.192.249***.259***.880.814 1.003.864 1.116.999.092***.076***.570**.365*** 1.327**.775*.992.937 2,581 2,170 2,584 2,171.343.379.225.283 526.60*** 535.58*** 48.90*** 192.85*** e From logistic and ordered logit regression procedures; each variable's second model includes all demographic variables as in Table 2, although the individual odds ratios are not shown. b Identity odds ratios are compared to theologically liberal Protestants. * p <.05 ** p <.01 *** p <. 001 scaled from one to three by level of participation, except for voting, which is dichotomous (0 = do not usually vote). When compared with Table 2, the playing field is more level here. Looking at each variable's first model, fundamentalists differ significantly from liberal Protestants in none of the actions. Mainliners' odds ratio of voting is twice that of liberals, but in protesting only about half that of liberals. The odds ratios for lobbying, voting, and religious community voluntarism for evangelicals are each significantly higher than those for liberals. The sole significantly higher odds ratio for practicing Catholics is in voting. Nominal Protestants appear significantly less active on three of the five activities, as do nonreligious persons. Weighing each variable equally, it appears that evangelical Protestants and practicing Catholics are the most active; fundamentalists and

1362 / Social Forces 76:4, June 1998 mainline Protestants are about as active as liberal Protestants; and the rest of the religious identities are less active. When demographic variables (not shown) and the attitude that religion is a public matter are included in the second models, as well as the interaction effect between evangelical and education, the significance of the original evangelical identity disappears.6 On none of the five dependent variables does the interaction effect appear significant. The original dependent variable that religion is a public matter, one that should address social and political issues bears significant impact on three of the public activities. It predicts higher levels of protesting and religious voluntarism, and lower levels of nonreligious community voluntarism. It has no significant effect on lobbying and voting. Discussion Conventional wisdom tells us that religion has retreated from the spheres of influence and halls of power now found in Hollywood studios, corporate boardrooms, and congressional committees. That may be the case. Nevertheless, privatization is multilevel, having institutional, organizational, and individual components (Chaves 1989, 1994; Yamane 1997). Our findings suggest that religion has certainly not become "invisible" or "Sheilaistic" at the individual level, and ostensibly likewise at the level of religious traditions (to quote Bellah et al. 1985 and Luckmann 1967). We have argued above that American religious deprivatization is a selective phenomenon, with some traditions privatizing for a time while others become more public and have offered historical and contemporary evidence of an almost cyclical pattern in American religious history. At different times, yet for religious reasons, conservative and liberal traditions (including Catholics) emerge for a time from a self-selected, privatized separation from public affairs to contest unacceptable public arrangements and legislation and to roll up their sleeves in religiously driven community voluntarism. Our findings support this theory in the current American religious field. There are wide and significant differences among religious traditions here. There is not, as some scholars have asserted (Marty 1970; Schmidt 1991), one public and one private party in American Protestantism, broadly referring as they were to liberal and evangelical Protestantism. Tables 2 and 3 refute this notion, pointing to the various positions on such a privatization spectrum that various Protestant, as well as Catholic, traditions inhabit. The results of our analyses point to conservatives, especially evangelical Protestants, as the religious Americans "going public" about their faith, constituting a clear departure from their frontier origins (Finke & Stark 1992), their historical premillennial theology, and their observed practice throughout much of the

Selective Deprivatization and American Religious Traditions / 1363 twentieth century. From Table 3, we can also assert that evangelicals' current deprivatization consists of more than just support for a religious legitimation of public involvement. While not as distinct from theologically liberal Protestants in actions as in attitude, they are nonetheless participating at levels no lower than the rest of religious Americans, and often higher. They lobby officials and participate in religiously based community organizations more than all other types. A reversal of great proportion and significance appears to be taking place, then, a "reversal" of the Great Reversal. Previous policies and theologies of noninvolvement in political and social issues appear to have given way to new forms of religious legitimation. And as Tables 1 and 2 (but not 3) attest, the inclination toward deprivatization is strongest among educated evangelicals. Holding a college degree actually strengthens their religious resolve for public engagement, reversing the effect that education typically yields.? Evangelical Protestants, and particularly the college-educated among them, are the most deprivatized religious Americans at present. Why is this the case, and how does this religious tradition turn such a traditional source of privatization (education) on its head? We suggest several reasons for this, including the "period effect" of Swidler's (1986) "unsettled lives" thesis, the adaptive fluidity of evangelical ideology, and the positive light in which evangelicals understand and appropriate higher education. First, religious conservatives as a whole appear to be currently deprivatizing, we have argued, in the wake of the unsettling era of the 1960s. Such a cultural shockwave required religious traditions to rethink the way they organize themselves, their priorities, and their interpretations of how they relate to "the world:' While liberal Protestants, and to a lesser extent mainliners, were supportive of the cultural changes that emerged from the 1960s, conservatives were not. And when confronted with the choice to be silent and adapt, or to speak out and resist, they appear to choose the latter option. Second, in the wake of such "unsettled" times and religious rethinking, it is important to bear in mind that conservative Protestant especially evangelical ideologies are historically fluid, a trait that all religious traditions exhibit to a greater or lesser degree. Evangelical theology is even considered dangerously malleable and innovative by some of its own scholars (Wells 1993). "Orthodox" biblical interpretation is not, nor ever has been, monolithic. Nor are evangelical religious rituals and symbolisms. For evangelicals, these traditions contain both privatizing and deprivatizing resources (Carpenter 1988; Hart 1992; Hunter 1983; Sikkink 1996; Smith 1998), "tools" in their "cultural tool kits" (Swidler 1986) that other religious traditions do not enjoy (or at least do not employ) to such an extent. Warner's (1988: 298) assertion that "religious Americans are not promising material for political mobilization," and that "evangelicals have an apolitical culture;' if once true, is now out of date. Biblical discourses about separation, purity, and the premillennial return of Christ (thus discouraging public activism) can and have

1364 / Social Forces 76:4, June 1998 been remolded to highlight public responsibility (Lienisch 1993). Evangelical and fundamentalist spokespersons have even recanted previous positions on public engagement, claiming misinterpretation of biblical texts (Lienisch 1993). Essentially, these religious cultures have new "strategies for action" (Swidler 1986). Third, fundamentalists have historically eyed higher education skeptically, convinced of its corrosive effects on religious faith and vitality (Marsden 1980). Evangelicals, however, have been less subject to such skepticism. Indeed, fundamentalist anti-intellectualism constituted one of the primary reasons that contemporary evangelicals broke ranks with them when the National Association of Evangelicals was formed in 1942 (Carpenter 1988; Henry 1947). Simply put, then, higher education composes for evangelicals a "cultural structure" that does not erode the conviction that personal religious faith ought to engage the public sphere and the broader culture. 8 It is one more "tool" in their distinctive cultural "tool kit" (Swidler 1986). Or, perhaps more accurately, education is a galvanizing coat, covering and strengthening their "tools" in a way it does not do for other religious Americans. Guth (1981), and to a more limited extent Schmalzbauer (1993), were correct in their assessments: modernizing forces have not only moved evangelicals into the middle class (a debate we do not examine here), but have raised their public consciousness and ire as they become newly equipped, by way of education, to mobilize on their ire. To sum up, in response to the unsettling period facing conservative Protestantism after the 1960s, new (and reversed) religious responses developed for engagement with a public sphere deemed hostile to their values a response not unlike that made by liberal Protestants in response to the Industrial Revolution. Among these theological conservatives, evangelicals have both ideological resources that legitimize involvement (since 1942) and a cultural structure in which education bolsters deprivatization instead of discouraging it. Whether they have maintained their "creek bank" values (Balmer 1996) is not for us to say, but their emergent deprivatization is supported here (as in Lienisch 1993; Schmalzbauer 1993). And it comprises more than a mere proclivity for voting (Kellstedt & Green 1993; Kellstedt & Smidt 1993; Wilcox, Jelen & Leege 1993), a reflection of status concerns (Gusfield 1963; Lipset & Raab 1981; Thurow 1996), or a result of lifecourse effects. Even denominational effects, biblical literalism, and charismatic identity do not temper the conservative religious identities' deprivatized attitudes. 9 Liberal Protestants and Roman Catholics have largely stood by while this reversal has taken place. This is partly explicable by way of the historically selective and cyclical nature of religious deprivatization outlined above. In addition, a "media effect;' implicating conservative Protestantism with all forms of religious activism, may stifle their willingness to admit religious legitimation for public engagement. Nevertheless, we do suggest that average liberal Protestants and Roman Catholics (and to a lesser extent mainline Protestants from results not shown) are genuinely not as interested in relating faith to public concerns as are their

Selective Deprivatization and American Religious Traditions / 1365 evangelical cousins. Yet from Table 3, it is evident that liberal Protestants and practicing Catholics are not inactive. But when interpreted in light of Table 2's results we conclude that the extent of their activism is lower than that of evangelicals and arises less from religiously inspired reasons than it does among evangelicals. It is likewise interesting to note that sheer belief in the merit of religious involvement in the public sphere predicts protesting and religious voluntarism independently of any identity. The extent of Catholic and liberal Protestant activism is also different in intent, to revisit Casanova's (1994) three reasons. 10 While our data is unable to affirm that Catholic and liberal Protestant activism accords with Casanovas second reason bringing moral and ethical considerations to bear on secular spheres of life we know that it is not about protection of traditional lifeworlds, Casanova's third reason. The sole exception lies with Catholic support for restricting abortion. Casanova's first reason is simply too vague and nearly inapplicable to the study of modern religious movements. Reasons that conservatives are currently deprivatizing are more evident. From their likelihood to vote Republican, oppose legal abortion, oppose easy divorce, and support Christian prayer in schools, evangelicals and fundamentalists "go public" for Casanova's third reason outlined above: to resist state (and we would add popular culture) penetration of the traditional lifeworld. Conservatives appear to be targeting select cultural structures, such as the family and education, to the neglect of other spheres (Calhoun 1992). This new Great Reversal, then, is not likely one of method or politics, for conservatives appear to have inherited few of the approaches associated with Social Gospel-style public engagement. Nevertheless, it would be false to suggest that evangelicals, Catholics, and liberal Protestants share no joint concerns. Data in Table 3 suggest that these three traditions share similar (insignificantly different) rates of participation in nonreligious voluntarism. As well, economic liberalism (Hart 1992) and commitment to poverty-relief organizations (Regnerus, Smith & Sikkink 1998) among evangelicals is higher than expected. Conclusion Theories of across-the-board privatization aside, there remains evidence of significant deprivatization at the individual level. A significant minority of Americans believe that their religious faith should address public issues of political and social significance and act accordingly. Among religious traditions, evangelical Protestants are the most resistant to the idea of a privatized religious faith, even when controlling for other religious and demographic factors. And while education fosters privatization in most people, it has the exact opposite effect on evangelicals, which we have sought to explain. Consistent with the idea that deprivatization is selective over time, liberal Protestants and Roman Catholics, once setting the

1366 / Social Forces 76:4, June 1998 benchmark for public religious belief and activism, are currently more private, suggesting evidence that a "reversal" of the Great Reversal has occurred. To a lesser degree, evangelical Protestants are also more publicly active Americans lobbying, voting, and volunteering in religious community organizations at higher rates than most. When controlling for demographics, the interaction effect between evangelical and education, and the original dependent variable, the odds ratios of participation in public activities by religious traditions become more equal. Our results are consistent with the historical cycle of selective deprivatization outlined herein. Public involvement, then, is not a stable feature of the differences between liberal and conservative Protestants, but one that appears to vary for reasons that are responsive to factors of historical context, available religiocultural resources, and the social backgrounds of people. Notes 1.Distinguishing between individual-level and organizational-level privatization is more difficult than between either of these and societal, institutional-level privatization, since the path between the former two levels, at least with regard to the formation of religious beliefs and the sustenance of religious plausibility through interpersonal networks, is both more evident and arguably much more important to the day-to-day maintenance of religious traditions than is the relationship between either of these levels and societal, institutional-level influences. Since this is the case, sociologists of religion (ourselves included) often make organizational-level assumptions and statements based on individual-level data. For example, strict churches (religious organizations) are strong, since individual religiosity among their adherents is strong. We do so as well in this article. 2. As Lincoln and Mamiya (1990) describe it, the black church in America has less frequently separated the sphere of the political from the sphere of the religious, despite their forced marginality from broader political culture. Nevertheless, our data constrains us from obtaining a clear portrait of the public or private status of the black church tradition, except as interpretation of the race variable allows, since the traditions we discuss here are of less value in understanding black Christianity. 3. Blau, Land, and Redding (1992) and Chaves (1989) provide evidence for significant "period effects" on religious adherence. After the Civil War, after World War I, and after the Vietnam conflict, religious adherence declined and then resurged (all significant coefficients except the post-civil War decline), perhaps reflecting the emergence of new religious strategies of action from older, less effective ones. Handy (1960) reports a similar dip-and-surge in adherence in his historical treatment of the American religious field after the Great Depression. 4. The "unsettled" era described here encompasses the sum of the ideological struggles marking the decade of the 1960s and its surrounding years. This includes American participation in Vietnam, the counterculture movement, the sexual revolution, civil rights, women's rights, and the fledgling gay and lesbian rights movement (see Epstein 1991).

Selective Deprivatization and American Religious Traditions / 1367 5.This neglect may be due to the univariate proxies so frequently utilized when studying religion. While there is a risk of generating multicollinearity among the covariates by using this approach, this neither justifies this method's exclusion nor constitutes a fatal flaw. Attentiveness to zero-order correlations and variance inflation factors can establish the presence of a serious problem. However, when we ran the identical regressions in OLS to account for this, no VIF exceeded 5.1, a reasonable range. None of the variables of most substantive interest in our results (evangelical, fundamentalist, etc.) displayed a VIF exceeding 2.2. In addition, a zero-order correlation matrix revealed that only one correlation among the four religious controls and the evangelical, fundamentalist, and liberal identities exceeded +/-.30 (that between conservative denomination and literal [.37]). This reinforces our conclusion that the religious variables we included are both theoretically and statistically distinct. 6. Demographic variables were included in Table 3 but do not appear for reasons of space and insignificant theoretical import. They are available from the authors upon request. 7. It is evident from Table 3 that the interaction effect between evangelicalism and education is not a factor in affecting odds ratios of participation in public actions, as it was in affecting the odds ratio of believing religion to be a public matter. This is neither surprising nor critical, for we are less concerned with how education predicts involvement than with how it shapes attitudes. The presence of a "public" attitude toward religion on the part of evangelicals is reason enough to suggest a reversal of the Great Reversal, actions aside. The attitude alone that religion is a public matter positively and significantly predicts protesting and religious voluntarism. And finally, unlike the phenomenon observed in the original dependent variable, education itself significantly predicts higher involvement in all five public activities except for protesting, where no significant relationship existed (odds ratios not shown). 8. Hunter's (1987) thesis in his study of evangelical college students that traditional orthodox belief is being eroded in the secularizing atmosphere of higher education is distinct from ours. We are concerned with the religious public engagement of evangelicals, not with measuring levels of traditional orthodoxy. Whether education is undermining theological orthodoxy or not, we cannot say. But education is equipping American evangelicals for public involvement in ways that it is not equipping other religious traditions. 9. Although less a central concern of this article, the positive significance of charismatic identity and biblical literalism in Table 2 ought not to be written off as tangential. Kellstedt and Green (1993), Kellstedt and Smidt (1993), and Wilcox, Jelen, and Leege (1993) were correct about their predictive importance. And Moberg (1977), while questioning what it would take for evangelicals to reverse the Great Reversal, correctly predicted that charismatics, frequently accused of sanctioning status quos, would be part of the process because of the confrontive nature of their religious worldview.

1368 / Social Forces 76:4, June 1998 APPENDIX A: Distribution of Identities by Biblical Literalism and Selected Protestant Denominations Fundamentalist Evangelical Mainline Liberal Other/DK Total % Strict Biblical Literalist 60.9 52.1 35.1 38.7 54.6 35.9 Denomination Total N "Fundamentalist" Southern Baptist 22.1 18.9 27.5 22.1 9.5 349 Miscellaneous 21.9 35.4 15.6 15.6 11.5 96 Pentecostala Independent Baptist 40.5 10.8 13.5 20.3 14.9 74 Assemblies of God 20.3 49.3 11.6 5.8 14.5 69 Missouri Synod 20.0 25.5 32.7 10.9 10.9 55 Lutheran Churches of Christ 34.6 13.5 19.2 13.5 19.2 52 Miscellaneous 37.5 31.3 12.5 15.6 3.1 32 Baptistb Missionary 14.3 7.1 14.3 28.6 35.7 28 Baptist National Baptist 7.4 18.5 22.2 37.0 14.8 27 (USA) AmericanBaptist 16.7 25.0 12.5 29.2 16.7 24 (USA) Evangelical Free/ 8.3 79.2 4.2.0 8.3 24 Covenant PCA 18.2 40.9 18.2 4.5 18.2 22 Free Will Baptist 47.1.0 23.5 23.5 5.9 17 CMA 14.3 57.1 7.1 14.3 7.1 14 Mennonite 21.4 28.6 7.1 7.1 35.7 14 Nazarene 16.7 33.3 16.7 8.3 25.0 12 Wesleyan Church 16.7 33.3 16.7 8.3 25.0 12 "Moderate" Disciples of Christ 29.0 3.2 51.6 9.7 6.5 31 American Baptist 26.7 10.0 30.0 26.7 6.7 30 Association Miscellaneous 5.0 15.0 65.0 10.0 5.0 20 Reformed' AME 26.7.0 20.0 40.0 13.3 15 AME (Zion) 8.3 25.0 8.3 41.7 16.7 12

Selective Deprivatization and American Religious Traditions / 1369 APPENDIX A: Distribution of Identities by Biblical Literalism and Selected Protestant Denominations (Continued) Denomination Fundamentalist Evangelical Mainline Liberal Other/DK Total N "Liberal" UnitedMethodist 9.9 10.3 48.8 20.7 10.3 242 ELCA 6.3 27.1 37.5 21.9 7.3 96 Presbyterian USA 4.8 11.1 54.0 23.8 6.3 63 Episcopal Church 4.8 7.9 49.2 28.6 7.9 63 United Church.0 3.4 51.7 41.4 3.4 29 of Christ Other Categories Nondenominational 29.7 31.9 14.5 16.7 7.2 138 "Just a Christian" 21.6 24.3 32.4 5.4 16.2 37 Note. For Biblical literalism, the table shows percent of total and, for selected Protestant denominations, it shows percent of total N for individual denomination.the denominational categories follow Smith's (1990) method of classification. a Includes Four Square Gospel, Full Gospel, Pentecostal Church of God, Pentecostal Holiness Church, United Pentecostal International, Church of God (Cleveland, Tenn. Apostolic Pentecostal, and "Other" Pentecostal. b Includes Primitive Baptist, Baptist Bible Fellowship, Baptist General Conference, Baptist Missionary Association, Conservative Baptist Association, Fundamentalist Baptist, and General Association Baptist. C Includes Christian Reformed Church, Reformed Church in America, and"other" Reformed. Notes (Continued) 10. Using other variables in the Religious Identity and Influence Survey, significant positive correlations exist between the evangelical and fundamentalist identities and variables coded as supporting laws against abortion, supporting spoken Christian prayers in schools, resisting the idea of divorce when a marriage becomes unhappy and unfulfilling, and voting Republican. No significant (positive or negative) correlations exist between liberal Protestants and any of these, except voting Republican (negative). Such lifeworld threats are simply not their battlegrounds. Our data (like most contemporary religious data) is less able to operationalize Casanova's (1994) second reason for deprivatization: bringing moral and ethical considerations to bear on secular spheres of life. Nevertheless, Demarath (1995) argues that the current decline in liberal Protestant public influence is due, ironically, to their success in establishing such values as freedom, tolerance, individualism, pluralism, and intellectual inquiry as acceptable in the broader culture. Their present privatization may be their reward for a job well done.

1370 / Social Forces 76:4, June 1998 References Ammerman, Nancy. 1982. "Operationalizing Evangelicalism: An Amendment." Sociological Analysis 43:170-71. Balmer, Randall.1996. Grant Us Courage: Travels along the Mainline of American Protestantism. Oxford University Press. Bellah, Robert N., Richard Madsen, William M. Sullivan, Ann Swidler, and Steven M. Tipton. 1985. Habits of the Heart. Individualism and Commitment in American Life. University of California Press. Berger, Peter L. 1967. The Sacred Canopy: Elements ofa Sociological Theory of Religion. Doubleday. Blau, Judith R., Kenneth C. Land, and Kent Redding. 1992. "The Expansion of Religious Affiliation: An Explanation of the Growth of Church Participation in the United States, 1850-1930:' Social Science Research 21:329-52. Calhoun, Craig. 1992. "Introduction: Habermas and the Public Sphere." Pp. 1-48 in Habermas and the Public Sphere, edited by Craig Calhoun. MIT Press. Carpenter, Joel A. (ed.). 1988. A New Evangelical Coalition: Early Documents of the National Association of Evangelicals. Garland. Casanova, Jos6.1994. Public Religions in the Modern World. University of Chicago Press. Chaves, Mark 1989. "Secularization and Religious Revival Evidence from U.S. Church Attendance Rates, 1972-1986?' Journal for the Scientific Study of Religion 28:464-77..1994. "Secularization as Declining Religious Authority." Social Forces 72:749-74. Demarath, N. Jay III. 1995. "Cultural Victory and Organizational Defeat in the Paradoxical Decline of Liberal Protestantism." Journal for the Scientific Study of Religion 34:458-70. Dixon, Richard, Lloyd Jones, and Roger Lowery. 1992. "Biblical Authority Questions: Two Choices in Identifying Conservative Christian Subcultures." Sociological Analysis 53:63-72. Dobbelaere, Karel 1981. "Secularization: A Multi-Dimensional Concept" Current Sociology 29:1-216. Epstein, Barbara. 1991. Political Protest and Cultural Revolution. University of California Press. Finke, Roger, and Rodney Stark. 1992. The Churching of America,1776-1990: Winners and Losers in Our Religious Economy. Rutgers University Press. Fowler, Robert Booth, and Allen D. Hertzke. 1995. Religion and Politics in America: Faith, Culture, and Strategic Choices. Westview. Greer, Bruce A., and Wade Clark Roof. 1992. "Desperately Seeking Sheila': Locating Religious Privatism in American Society." Journal for the Scientific Study of Religion 31:346-52. Gusfield, Joseph. 1963. Symbolic Crusade: Status Politics and the American Temperance Movement. University of Illinois Press. Guth, James. 1981. "The Politics of the Evangelical Right: An Interpretive Essay." Paper presented at the annual meeting of the American Political Science Association. Guth, James L., and John C. Green (eds.). 1991. The Bible and the Ballot Box: Religion and Politics in the 1988 Election. Westview Press. Handy, Robert. 1960. "The American Religious Depression, 1925-1935:' Church History 29:2-16.

Selective Deprivatization and American Religious Traditions / 1371. 1984. Christian America: Protestant Hopes and Realities. Oxford University Press. Hart, Stephen. 1992. What Does the Lord Require? How American Christians Think about Economics. Oxford University Press. Henry, Carl F.H. 1947. The Uneasy Conscience of Modern Fundamentalism. Eerdmans. Hunter, James Davison. 1983. American Evangelicalism: Conservative Religion and the Quandary of Modernity. Rutgers University Press..1987. Evangelicalism: The Coming Generation. University of Chicago Press. Kellstedt, Lyman A., and John C. Green. 1993. "Knowing God's Many People: Denominational Preference and Political Behavior." Pp. 53-71 in Rediscovering the Religious Factor in American Politics, edited by David C. Leege and Lyman A. Kellstedt. M.E. Sharpe. Kellstedt, Lyman A., and Corwin E. Smidt. 1993. "Doctrinal Beliefs and Political Behavior: Views of the Bible." Pp. 177-98 in Rediscovering the Religious Factor in American Politics, edited by David C. Leege and Lyman A. Kellstedt. M.E. Sharpe. Lienesch, Michael. 1993. Redeeming America: Piety and Politics in the New Christian Right. University of North Carolina Press. Lincoln, C. Eric, and Lawrence H. Mamiya. 1990. The Black Church in the African-American Experience. Duke University Press. Lipset, Seymour M., and Earl Raab. 1981. "The Election and the Evangelicals." Commentary 71:25-32. Lovin, Robin W.1986. "Religion and American Public Life: Three Relationships." Pp. 7-28 in Religion and American Public Life: Interpretations and Explorations, edited by Robin W. Lovin. Paulist Press. Luckmann, Thomas. 1967. The Invisible Religion: The Problem of Religion in Modern Society. Macmillan. Luker, Kristin. 1984. Abortion and the Politics of Motherhood. University of California Press. Marsden, George M. 1980. Fundamentalism and American Culture: The Shaping of Twentieth- Century Evangelicalism 1870-1 925. Oxford University Press.. 1994. The Soul of the American University: From Protestant Establishment to Established Nonbelief. Oxford University Press. Marty, Martin E. 1970. Righteous Empire: The Protestant Experience in America. Dial Press. McAdam, Doug. 1988. Freedom Summer. Oxford University Press. Moberg, David O. 1977. The Great Reversal. J.D. Lippincott. Regnerus, Mark D., Christian Smith, and David Sikkink. 1998. "Who Gives to the Poor? The Role of Religious Tradition and Political Location on the Personal Generosity of Americans toward the Poor." Journal of the Scientific Study of Religion 37:3. Schmalzbauer, John. 1993. "Evangelicals in the New Class: Class versus Subcultural Predictors of Ideology." Journal for the Scientific Study of Religion 32:330-42. Schmidt, Jean Miller. 1991. Souls and the Social Order: The Two-Party System in American Protestantism. Carlson. Sewell, William H., Jr. 1992. "A Theory of Structure: Duality, Agency, and Transformation." American Journal of Sociology 98:1-29. Sherkat, Darren E., and T. Jean Blocker. 1997. "Explaining the Political and Personal Consequences of Protest" Social Forces 75:1049-76.

1372 / Social Forces 76:4, June 1998 Sikkink, David. 1996. "Saving Kids, Saving Schools: Evangelicals and Schooling Strategies for Children.' Paper presented at the 1996 annual meeting of the Association for the Sociology of Religion. Smith, Christian. 1996a. "Correcting a Curious Neglect, or Bringing Religion Back In." Pp. 1-25 in Disruptive Religion: The Force of Faith in Social MovementActivism, edited by Christian Smith. Routledge.. 1996b. Resisting Reagan: The U.S. Central America Peace Movement. University of Chicago Press..1998. American Evangelicalism: Embattled and Thriving. University of Chicago Press. Smith, Tom. 1990. "Classifying Protestant Denominations." Review of Religious Research 31:224-45. Swidler, Ann. 1986. "Culture in Action: Symbols and Strategies." American Sociological Review 51:273-86. Thurow, Lester. 1996. The Future of Capitalism: How Today's Economic Forces Shape Tomorrow's World. William Morrow. Warner, R. Stephen. 1988. New Wine in Old Wineskins: Evangelicals and Liberals in a Small- Town Church. University of California Press. Weber, Max. 1986. Economy and Society, edited by Guenther Roth and Claus Wittich. University of California Press. Wells, David F. 1993. No Place for Truth, or Whatever Happened to Evangelical Theology? Eerdmans. White, Ronald C., Jr., and C. Howard Hopkins. 1976. The Social Gospel: Religion and Reform in ChangingAmerica. Temple University Press. Wilcox, Clyde. 1989. "The New Christian Right and the Mobilization of the Evangelicals" Pp. 139-56 in Religion and Political Behavior in the United States, edited by Ted G. Jelen. Praeger. Wilcox, Clyde, Ted G. Jelen, and David C. Leege. 1993. "Religious Group Identifications: Toward a Cognitive Theory of Religious Mobilization:' Pp. 72-99 in Rediscovering the Religious Factor in American Politics, edited by David C. Leege and Lyman A. Kellstedt. M.E. Sharpe. Williams, Rhys H., and N.J. Demarath III.1991. "Religion and Political Process in an American City." American Sociological Review 56:417-3 1. Wuthnow, Robert. 1983. "The Political Rebirth of American Evangelicals." Pp. 167-85 in The New Christian Right: Mobilization and Legitimation, edited by Robert C. Liebman and Robert Wuthnow. Aldine.. 1987. Meaning and Moral Order: Explorations in Cultural Analysis. University of California Press. 1988. The Restructuring of American Religion. Princeton University Press. 1989a. Communities of Discourse: Ideology and Social Structure in the Reformation, the Enlightenment, and European Socialism. Harvard University Press..1989b. The Struggle foramerica's Soul: Evangelicals, Liberals, and Secularism. Eerdmans. Yamane, David. 1997. "Secularization on Trial: In Defense of a Neosecularization Paradigm." Journal for the Scientific Study of Religion 36:109-22.