Advance contract signed June 2000, Princeton Series in Cultural Sociology, edited by Paul DiMaggio, Michele Lamont, Robert Wuthnow, and Viviana Zelizer, Princeton University Press.

In the early 1990s, while conducting research for a study of conflict and decision-making in local congregations, I interviewed about 250 people in and around Oak Park, Illinois. A medium-sized suburb on the edge of the city of Chicago, it was and is predominantly a community of young, professional families. When I walked into the churches, parish halls, and synagogues of this community, I would often see an old black-and-white photograph prominently displayed in the foyer. The pictures all had a similarity I did not really think much about at the time. Row upon row of children were lined up, usually outside in the sunshine, all dressed up in the kinds of dresses or shirts and ties that were the proper attire for the children of the day. These were the overflowing religious education classes of the 1950s and early 1960s, when most white suburban middle-class children in places like Oak Park grew up in male-breadwinner families, and church attendance reached a century-high peak.

In each interview with pastors and lay leaders, I asked about the most pressing challenges facing their congregations over the next five to ten years. Most comments centered around the typical organizational concerns all congregations face: how to attract new members and manage growth, how to increase the budget, the hiring of a new pastor. But other than that, by far the most common theme was how to develop ministries that would be relevant in people's lives in the face of changes in family and work.

Some congregations were struggling to find forms of women's ministry that would continue to meet the needs of their increasingly smaller portion of homemakers while also attracting younger "working" women. In other congregations, members were calling for ministries for single parents, working parents, or those going through divorce or recovering from addiction. In some congregations, gay and lesbian members were asking for some kind of religious ritual that would acknowledge their unions as "families," too. A community of geographically mobile professionals, many Oak Park residents were turning to congregations to be a kind of surrogate extended family, meeting not only religious, but also social and emotional, needs for the entire family. Some congregations had already responded to these changes with enthusiasm, organizing new ministries or incorporating new kinds of people into more traditional groups for women, men, or parents. But many congregations were just beginning to recognize these issues, and were unsure of how to formulate a response.

This surprised me very much at the time. Women's rates of labor force participation had been rising for more than two decades. Divorce, cohabitation, and other disruptions of a typical 1950s family formation cycle had been on the rise for a long time. I began to wonder why so many congregations were only now responding to changes in work and family. And what effect were these changes having on congregational membership, programs, and discourse about what constitutes a good or morally correct family lifestyle? It did not take very much time for me to discover that, while understanding these social changes has been part of our scholarly discourse for close to two decades, there is very little research on their impact on the life of local community institutions, especially religious congregations.

As a result, I decided to begin a new study that would analyze the effect post-1950s changes in work and family life have had on religious institutions, and how religious institutions fit - or do not fit - into contemporary lifestyles. Is participation in local congregations still tightly linked to family formation? What effect does the "time squeeze" associated with the dual-earner lifestyle have on religious involvement? In a period of declining numbers of two-parent-with-children families, and very few male-breadwinner families, who feels included or excluded in local congregations? What ideas about "the good family" are being shaped by local religious cultures and practices?

In formulating a way to conduct research that would shed light on these questions, I made several key choices. First, I decided that such a study should be conducted in a way that would allow for simultaneously answering two kinds of questions. What has happened inside religious organizations and institutions faced with changes in work and family? And how do religious beliefs, practices, and institutions affect the lives of individuals and families in contemporary society? I decided the study must be an analysis of reciprocal relationships, examining how the relationship between two historically connected institutions has changed over time.

I also made decisions about the type of study it should be, about the research design and methods. I decided the study should be conducted at the level of the local congregation. Religious elites make pronouncements about family life, and the scholarly discourse on religious groups' family ideologies is quite well developed. But local congregations directly provide ministry to individuals and families, and it is these organizations in which the vast majority of the religiously involved participate. Unlike some other, newer forms of local religious community, congregations were associated in the 1950s with a very specific work-family lifestyle, and so changes in work and family might pose particular challenges for them. Local congregations may fall in line with official pronouncements, but in fact they often do not, particularly in areas like women's roles or lifestyle choices. Yet our knowledge of how the religion-family link works at the level of the local congregation is limited to a few case studies, a high percentage of which focus on evangelical Protestants alone.

Another choice was to take an explicitly ecological approach to the research design. An ecological approach implies a focus on a few specific communities in order to understand how the local context influences processes of social change. Particularly important was the decision to pick communities that varied in socio-economic status and in population density. Many studies of American religion have focused on a white, suburban middle-class environment, and while infinite variation is not possible in a small-scale study, some variation in resources and lifestyles is desirable because it prevents universalizing one particular set of middle-class experiences as "the" American religious experience.

An ecological approach also allows for an examination of how innovations in ministry spread through local mechanisms like pastor's networks, or result from local ecological processes, like pastors borrowing programs of successful peers or competitors. This implies a very different methodology than, say, a national random-sample survey of pastors with a few contextual variables included.

This book reports on the findings of the Religion and Family Project, which analyzes the religion-family nexus in four upstate New York communities, chosen to vary by socio-economic status (middle-and working-class) and by population density (roughly, "rural" and small-town contexts versus urban contexts). Over two years in 1998 and 1999 I collected data in these communities through a survey of pastors, a survey of community residents, participant-observation in local congregations, and follow-up in-depth interviews with a small number of survey respondents.

In upstate New York in the late 1990s, I found many issues similar to the ones named by pastors and congregation members in Oak Park several years ago. I also found new issues and themes emerging. This book examines these themes. One of the major goals is to provide a good map of the fit between local congregations and contemporary work and family lifestyles in these four communities. The other goal is to understand the implications of this new map for our theoretical accounts of the nature and outcome of religious institutional change in the United States.

Throughout the history of the United States, religion and family have been two institutions linked together through relationships of dependency and control. Mainstream religious institutions have responded to evolving family life and household arrangements by providing new ministries to meet families' needs in an ever-changing social context. New religious groups and movements like the Shakers have formed around alternative family and household arrangements. But if changes in family life have had an impact on how religious institutions are organized, religious institutions have also provided moral guidelines that shape family practices, the organization of family life, and our perceptions of "the good family."

Religious institutions have depended on families to pass on the rituals and beliefs of a particular faith tradition. In colonial times, fathers, as the head of an extended and production-oriented household, were often primarily responsible for the religious instruction of their wives and children, along with other dependents. In the nineteenth century an industrial economy fostered more nuclear-family households, and in the latter half of that century an emerging urban middle-class with a male-breadwinner lifestyle. A new emphasis on mothers' responsibility for the religious well-being of children and husbands emerged in the context of home-centered religious education and devotional practices.

In the late nineteenth and early twentieth centuries, Sunday School and other formal religious education programs were partly successful in moving the religious instruction of children out of the home and into the congregation or parish; these did not eliminate, but rather worked to reinforce, the home-based rituals central to religious socialization. McDannell (1995) argues that in the Catholic church, there began a century-long struggle over the "feminization" of religion, with religious elites working to retain religious authority in the face of home-centered religious practices that often deviated from officially-endorsed ritual and doctrine. A similar tug-of-war between lay women and mostly male religious elites had been occurring for some time in Protestant traditions. This tension became heightened in post-WWII America, as Protestant and Catholic churches thrived in the rapidly-expanding middle-class suburbs. As local congregations organized ministries for women and children, religious leaders warned anew against the "feminization" and "privatization" of religion.

If families have influenced religious institutions, they have also depended upon them for the moral socialization of children, for reinforcing a sense of ethnic identity and continuity, and for important rituals that mark life-course transitions - weddings and funerals, bar and bat mitzvahs, baptisms and brises. Confirmation and Sunday School classes welcome children into moral adulthood; churches pass out Bibles with students' names stamped in gold letters on the front to seniors in celebration of high school graduation; priests bless the celebration of girls' "quince años" in Mexican-American communities. Religious involvement, many argue, has been functional for families, associated with lower rates of divorce and higher levels of satisfaction in marriage, parenting, and other family relationships.

These relationships of mutual dependency have also had an aspect of social control. Religious institutions in the United States have promoted familism, the ideology that the family is the central, most fundamental unit of social order in a society. Familistic ideologies are historically specific, and vary over time and place. They do not promote the idea that any kind of family is equally valid, but rather tend to idealize certain forms and functions of the family, defining them as legitimate, valuable, and morally correct, even essential for a healthy social order.

A few utopian religious communities have experimented with forms of communal organization that do not rely on some form of nuclear, patriarchal family as the basic unit of social organization. But generally, mainstream religious institutions in the United States have promoted stable, monogamous marriages which produce children. They have bolstered parental authority, and have discouraged pre-marital and extra-marital sex. Religious institutions have contributed to a normative consensus in our society, stronger in some times and places than in others, that being an unmarried adult, deciding to remain childless, or living in a same-sex union are at best unfortunate states in which to find oneself, and at worst irresponsible or deviant choices which should be sanctioned.

For much of American history, religious institutions have promoted ideologies that interpret men's and women's natures as fundamentally different, and they have encouraged the development of various versions of the ideology of separate spheres, with male activity concentrated in the realm of work and civic life (defined as public) and women's activity concentrated in the home and church (defined as private). Many mainstream religious groups today endorse various forms of traditionally gendered family roles.

Of course, the specific forms of family behavior and specific ideals of the good family have varied over time and according to ethnicity, social class, and religious tradition. Churches in the African-American tradition have found ways to value extended, maternally-based family networks, although they have also sought to bolster male familial authority in the context of a society that systematically disempowers black men. In the 1960s, sociologists documented large Catholic/Protestant differences in childrearing and other family behaviors, with Catholics favoring more extended family networks and emphasizing the importance of obedience and conformity in children. In the 1980s Alwin (1986) argued that such differences had largely disappeared as Catholics experienced upward social and economic mobility. He found that the major differences in childrearing norms and behaviors were between those who attended church (any church) and those who did not. Church attenders were more likely to favor conformity in children and traditional gender roles.

In the 1990s, sociologists began to argue that the major differences in family ideology and behavior today are driven by the presence of a distinct conservative Protestant subculture, which endorses an ideology of male household headship and a preference for obedience in children. Bendroth (1993) however, shows that the label "conservative" is too broad to characterize the response of those on the religious right to post-50s family change. She argues that fundamentalists have resisted any rhetoric that embraces equality between men and women and have maintained a strict focus on paternal authority and obedience in children in practice, as well as in rhetoric. Evangelicals, while maintaining a rhetoric of the man as "spiritual head" of the house, have been more flexible and adaptable on questions of women's labor-force participation and have remained more egalitarian and nurturing in practice than in rhetoric.

Such religiously based family norms affect behaviors within families, but some have also become overtly politicized. One of the major grounds of the "culture war" has been family norms and behaviors. Lakoff (1996) believes that two different cultural models of the family are linked not only to liberal/conservative differences in the areas of birth control, abortion, gay marriage, and other "family values" issues, but also underlie many other culture war policy differences, including different ideas about economic development, welfare, and access to political representation. Because liberal and conservative models of the family are based upon and uphold fundamentally different understandings of the division between the public and the private, they foster different understandings of the link between family behavior and citizenship.

Lakoff (1996) raises an important point. Many scholars have treated religion and family as "private" institutions. But they are both, in some important aspects, public institutions as well. Families link the individual to broader connections with the workplace, with civil society, and the state. The way in which families are defined, then, can have important public implications. To take just one example, comparable worth policies may find less support among those who endorse a religiously-based view of the man as the family breadwinner. Feminists have long argued that families are not naturally "private" but are defined as such, in part by religious institutions. Defining family as "private" systematically de-emphasizes the importance of women's typical daily lives and concerns in the political arena. As a result, women are not in practice treated as full citizens in the same sense as are men, and they are vulnerable to the violation of their rights, including in some cases the very basic right to physical safety within their homes.

Religious institutions are also public in the sense of being a part of civil society not controlled by the state. They provide a public arena of moral discourse and a location for building "community" and the social capital that entails. Religious institutions provide ideologies and identities that mobilize people for political and social action on a wide range of issues. The role of religious leaders in the controversy over gay marriage is a concrete example of how religious institutions mediate the relationship between the family and the state, in this case providing a moral rationale for laws that designate heterosexual marriage as a specific kind of legal contract with publicly-enforceable obligations, while not providing that status for same-sex unions.

The assumption that religion and family are private institutions has dominated scholarship in the sociology of religion, but it seems clear that these institutions operate in both the public and the private sphere as these traditionally have been understood. It seems more analytically useful, then, to conceptualize both religion and family as institutions that connect individuals to public life, and to ask questions about how those connections are conceptualized and practiced in different times and places. Three kinds of questions seem especially helpful in designating a location from which to begin a different kind of scholarly inquiry into the religion-family link. Do religious institutions today define the family, in practice and in rhetoric, as a private institution? What cultural model of the family are contemporary congregations organized around? And does the historical association between religious institutions and a privatistic, "separate spheres" family ideology make them a less comfortable fit for contemporary men and women, who have experienced within one or two generations rapid changes in work and family roles?

These questions become even more central because the last great religious expansion in the United States institutionalized programs and ways of doing things that facilitated a revival of a form of familism that defined the family as a private "haven in a heartless world," and which assumed a traditionally gendered division of labor in the home. The 1950s saw rapid growth in organized religion and century-high levels of church attendance, as new congregations sprang up overnight, especially in the booming post-war suburbs. This 1950s religious revival was driven by changes in commuting patterns, rising marriage and birth rates, changing child-rearing patterns, and the prosperity which allowed more and more families to adopt the male-breadwinner lifestyle.

Congregations embraced these developments with enthusiasm, expanding their programs for children, teens, and women, and developing social activities for the whole family. Such programs became institutionalized as a "standard package" of family ministries. This standard package not only assumed a two-parent family, but was based upon a specific middle-class male-breadwinner version of it. Even though this lifestyle was not available to all congregation members, it was upheld as an ideal, and congregational practices and rhetoric were organized around supporting this the middle-class, suburban, organization-man lifestyle.

For example, the increasing numbers of women's ministries generally met during the daytime. This made them largely unavailable to women who worked outside the home. But for stay-at-home mothers, daytime United Methodist Women meetings (or the Altar Guild at the Catholic parish or the Sisters group at the Reform temple) meant evenings that were free for spending time with their families or watching the children while husbands went to evening church (parish, temple) council meetings.

The male-breadwinner lifestyle must have seemed not only right but even inevitable to many church leaders in the 1950s. The proportion of households comprised of two married parents and one or more children peaked, reaching a century-high level of 43% with the post-War baby boom. But what may have seemed like historical inevitability turned out to be only an anomaly, and a relatively short-lived one at that. By the 1990s the male-breadwinner family was only a memory to most Americans, surviving in a few of the most affluent suburban enclaves. Congregations found themselves facing an entirely different landscape.

In the 1990s, family life is very different than it was in the 1950s for the vast majority of Americans. Table 1 shows some of these differences at a glance. The dual-earner family has become statistically dominant. Skolnick (1991) argues that the rise of the dual-earner couple has not set a single new norm for family life but has led, along with other social and cultural changes, to an increasing cultural pluralism in the area of the family. Many people have a more flexible image of what families are and should be like, with less emphasis on a single normative family form. Some groups in society do hold very specific views of what is normative in family life, but these norms differ across different groups. However, a large and growing literature on the dual-earner couple supports the idea that a specific dual-earner lifestyle, based primarily on the norms and values of middle-class managers and professionals, has in fact become culturally normative in our society, and that other institutions are beginning to adapt to this lifestyle.

[TABLE 1 ABOUT HERE]

Of course, the dual-earner couple is not a new family form. In 1950, both parents spent at least some time each year in the paid labor force in fully half of all families with children. But the dual-earner family has become much more common, and the number of joint hours spent at paid work by husbands and wives has rapidly increased. And the dual-earner family form has spread widely throughout white, middle-class America, a marked change from the 1950s. Most mothers, even of very young children, now work outside the home for pay, many of them working full time. The changing relationship between work and family is one of the most fundamental transformations in our society.

There has been an ongoing debate about the effects of this transformation on family life. Some argue that the "time squeeze" encountered in dual-earner households has caused greater levels of stress and anxiety for women and various other problems in managing family life on a daily basis. Others argue that women who work outside the home have less stress and better relationships with their spouses and children, and point to the beneficial effects of having two jobs to buffer the family from economic uncertainty. Nock (1998) argues that the rise of the dual-earner couple has led to fathers who are more involved in the parenting of their children and more involved in the home in other ways, as well. More generally, if Americans work more, they are also getting more education and working better jobs for higher pay.

In addition to the rise of the dual-earner household, there are more and more households not organized around two parents raising children. Much larger and more stable portions of the population remain unmarried, or childless, throughout much of their adult lives than in the 1950s. More couples live together, and raise children, without being married. There are more single parents; divorce and blended families are much more common. Gay and lesbian lifestyles, while still controversial in some parts of American society, have gained a great deal of legitimacy and visibility. As Furstenberg notes, "Marriage is no longer the master event that orchestrates the onset of sexual relations, parenthood, the departure from home, or even the establishment of a household. These events have become . . .discrete moment in the life course."

Accompanying the changes in family form and in women's roles in society have been changes in people's attitudes and beliefs about gender roles. Nock (1998) profiles an increasing importance of marriage and children in men's lives with the move away from a male-breadwinner society. Prior to the 1970s, polls show that most Americans agreed with statements like, "It's better for everybody if the man earns the money and the woman takes care of the home and family." Since the mid-70s, Americans have displayed more egalitarian ideas about gender roles, and have come to favor not only fathers' increased involvement in childrearing, but also policies that grant women greater access to "public" goods - good jobs, equal pay, and political power. More generally, Americans have come to have an individualistic orientation toward many institutions, including work, family, and religion, and to feel that assigning individuals roles in society based on ascriptive characteristics - gender and race, for example - is wrong.

As a result of these changes, mainstream religious institutions in the United States have faced a large-scale and fundamental transformation in the institution to which their fate historically has been most closely tied. A period of rapid religious expansion and institution-building in the 1950s was organized around a form of family life that quickly peaked and rapidly faded away. It is not that all families in the 1950s were male-breadwinner; fully half were not. It is not that suburbs, or even male-breadwinner families, have completely disappeared. But religious organizations that promoted a suburban male-breadwinner lifestyle as ideal and organized themselves around this model of the family found themselves with programs, rituals, and discourses targeted to a rapidly decreasing proportion of the population. How have religious institutions responded? And has the response led to the maintenance of the tight link between religion and family that has been characteristic over most of the course of our history?

There has been relatively little work analyzing how religious institutions have responded to these changes in work and family. In a case study of one mainline Protestant church she calls Briarglen, Marler (1995) outlines what she believes may be a typical response of churches within that tradition. Briarglen has been successful in attracting a younger generation of dual-earner families with children. These families have less time and more money than the previous generation of members, and their participation largely takes the form of writing checks to support ministries for their children, while an older generation provides the labor to make the ministries work.

Marler (1995) is critical of this pattern for several reasons. By organizing ministry around two-parent families with children, she argues that this church is exhibiting a "nostalgia" for the 1950s, and ignoring the increasing numbers of single parents and long-term singles in the population, thus limiting their target market. She also argues that the division of labor at Briarglen takes advantage of the older generation who ought to be able to relax and enjoy their lives, instead of providing services for others. And she is concerned that this pattern indicates a decline in religious commitment among younger cohorts. Marler asks whether the current generation of dual-earner parents will bother to stay with the church once their children are grown, and she raises the question of whether the mainline Protestant decline in membership since the 1960s may be linked to an inability to form a better response to changes in women's roles, work, and family.

Other research indicates that the pattern Marler (1995) discovered at Briarglen may be more positive, especially in its consequences for the older generation. For those at or approaching retirement, volunteer activity has been consistently linked with increased health and lower rates of depression in later life. As for the other negative consequences that Marler writes about, it is difficult to judge how much of a problem they really present, since there are no good comparative organizational studies of how congregations have responded to these social changes.

But other evidence suggests that liberal and moderate Protestant churches may have resisted innovating to meet the needs of those who do not fit a two-parent-with-children family pattern. Fishburn (1991) argues this is directly linked to the decline of mainline Protestant churches since the 1960s. In general, leaders in these traditions have a strong preference for putting energy to develop new ministries into more valued "public" outreach - peace, justice, and poverty ministries. A focus on family ministry is often seen as directly competing with more "public" forms of outreach for members' time, money, and efforts. Browning (1997) provides support for this interpretation when he calls for members of mainstream religious groups to rejoin the public dialog on family issues which has largely been ceded to evangelical Protestants, who see the family as a central focus of mission, not a distraction from more important goals or activities.

As Lakoff (1996) notes, conservatives have not had a similar hesitancy in making family ministry - and family values - a central part of their rhetoric and programming. Evangelical leaders have responded to the increasing gender egalitarianism of American society by developing discourses that emphasize mutuality, caring, and sharing within the household rather than a strict exercise of patriarchal authority. A host of later research confirms Bendroth's argument about evangelical Protestants having fostered more egalitarianism in practice than their rhetoric would imply. Demmitt (1992) and Bartkowski (1997) find that evangelical Protestant congregations have adapted well to dual-earner families. In a previous study (Becker 1999), I found that both Catholic and evangelical Protestant pastors saw the development of more contemporary forms of family ministry as a natural extension of their prior activities, and a positive mission priority.

Although we have focused a great deal of attention on the official responses of religious leaders to changes in work, family, and gender, looking at the rhetoric developed in advice books or other "official" texts, our knowledge about congregational response to family change is very piecemeal. There have been a few case studies of individual congregations. But we have no good comparative studies of local congregations that would allow us to identify how and why they vary in the kinds of family ministry they provide, or the variation in the kinds of individuals and families that they attract. So it is hard to know if Marler's study of Briarglen or Demmitt's study of an evangelical congregation are typical and what the range of response is within and across religious traditions. We know relatively more about evangelical Protestant congregations, but have little knowledge of what is going on within Catholic parishes, Jewish synagogues, or other religious communities.

In addition, we need to understand the impact of religious institutions' adaptations on their ability to attract contemporary families and individuals. Do those who are not in a male-breadwinner or intact two-parent family arrangement find religious organizations relevant and meaningful in their lives? Do they feel welcome in local congregations? What effects does religious involvement have on individual's choices about work and family, or on behaviors within families from childrearing to relationships between husbands and wives? Constructing a new map of the religion-family nexus requires understanding not only what religious institutions are doing, but also how religious institutions are linked to family formation and work for the contemporary families - and individuals - of the 1990s.

In terms of understanding the effect of changes in family and work on religious participation, scholars have looked at only a limited set of questions. There has been little research on the effects of divorce and other forms of family disruption on religious involvement, and little research on the involvement of single parents, those in blended families, or long-term singles. We know that family formation is still associated with religious involvement, but we do not know if this is equally true for men and women, since research on men's and women's involvement has focused almost exclusively on the "gender gap" in church attention, and not on the processes of religious affiliation or the meaning religious involvement has in men's and women's lives. We know that paid employment increases men's religious involvement, but there has been little research on how religious involvement might affect individual's decisions about the hours they spend at work or about accepting promotions or transfers to jobs with more travel.

If there are gaps in what we know about congregations and the religious commitment of individuals, there are also silences created by our conceptual categories. Most sociologists accept at face value that idea that religious involvement is functional for families, and this has consequences both for the interpretation of research findings and for what kinds of questions are asked (Becker 2000). For example, scholars emphasize not the structural inequality between men and women in evangelical churches, but the negotiated and more egalitarian practices within evangelical families. We know from other research that work-family issues are treated largely as private matters in the United States, and that businesses off-load the costs of managing work and family onto individuals, thus "privatizing" the cost of coping with social change (Becker and Moen 1999). Do religious institutions contribute to this idea that family is a private institution and that troubles managing work and family are private troubles, the costs of which should be born by individuals, not companies or the state? There has been little or no research on questions such as these.

One of the main goals of this book is to understand how congregations have adapted to changes in work and family in their discourse, in their formal programming, and in their everyday practices. Another goal is to analyze how individuals' work and family context affects religious involvement in a way that includes an examination of family disruption, alternative life-course pathways, and non-traditional family arrangements. This will generate a good map of the current linkages between religion and family in these four communities. But together, the answers to the questions posed above also add up to a larger pattern that can help us to refine our theoretical account of how religious institutional change comes about. This more theoretical analysis will allow for a critical examination of theories about the role of religion in the contemporary United States, as well as theories about the overall direction or outcome of religious institutional change.

In developing a theoretical understanding of how changes in work and family have affected religious institutions, there are two kinds of questions that need to be addressed. The first set of questions has to do with the processes through which change occurs. How have religious institutions adapted to changes in work and family? What are the processes and mechanisms through which religious institutions define some changes as relevant, decide on a response, and carry out that response? In the sociology of religion, processes of religious institutional change have been analyzed using two different theoretical frameworks. The first is a market framework, based on rational choice theory. The second is an institutional framework drawn from organizational theory and the sociology of culture. These frameworks are not unique to the sociology of religion, but have been developed and applied in an effort to understand a wide variety of social processes across institutional realms.

The second set of questions has to do with the overall result of the changes that have come about. Are religious institutions still as tightly organized around family formation and family life as they were in the 1950s? Do they still have the same impact on family life that they did in the 1950s? What are the consequences of the changes in family and work for the ongoing vitality of religious institutions in American society? The outcomes of religious institutional change have been understood through frameworks that are more centrally located within sociology of religion as a distinct discourse. The first framework is that of secularization theory, which has developed from the tradition of sociological theory inspired by the work of Max Weber. The second framework draws upon the work of the French sociologist Pierre Bourdieu and conceptualizes religious institutional change as a process of defining and defending an evolving orthodoxy.

The idea of a religious "market," complete with consumers and entrepreneurs, borrows from rational choice theory and classical organizational theory in sociology. This framework understands individuals as consumers of religious goods and services, who make rational choices (choices that maximize the benefits and minimize the costs) about participating in religious institutions that meet their socially-influenced preferences. How religious preferences are formed has received quite a bit of attention in the rational choice approach to religion, and is often linked to family relationships. For example, religious socialization into a particular tradition as a child makes it familiar and lowers the costs of participation, leading to stability in religious preferences. As an adult, a spouse who attends a particular church may exert either a normative influence on one's own choice to attend, or provide information that makes the given organization more familiar or more trusted, reducing the learning costs associated with participation.

This perspective lends itself to certain kinds of predictions about how changes in work and family might affect religious participation. To take just one example, if there has been a loosening of the connection between family formation and religious socialization, this might lead to an overall decline of religious participation rates in the population. Or, it might lead to a reconfiguration of the religious market, with those organizations that do a better job of retaining young families expanding to occupy a greater market share, competing effectively in the battle to retain future generations of adherents. In fact, some rational choice proponents have made this argument to explain the processes through which mainline churches have declined and evangelical Protestant churches have thrived over the last three decades.

Like changes in family formation, changes in work might reduce religious participation or merely reconfigure it. The long hours spent at work by managers and professionals caught up in the "time squeeze" makes less time available for religious activities, and paid employment may provide women, who have typically been more religiously involved than men, with alternate venues for achieving status and other rewards they previously received from religious organizations, thus reducing religious involvement. On the other hand, early religious socialization might counteract these tendencies for some individuals. And those with lower occupational attainment might turn to religious groups for status and other forms of affirmation they do not receive at work.

From this perspective, religious organizations also make rational choices to react to changes in the market for religious goods and services. Or at least, religious leaders do; most rational choice perspectives emphasize the role of religious "entrepreneurs" who are relatively free, given the largely congregational structure of American religion, to innovate in ways that attract new adherents by providing services that religious consumers really want. This leads, proponents of this framework argue, to a vital and pluralistic religious landscape, that provides an overall high level of quality and diversity in religious goods and services. This makes it even more likely that changes in work and family will not lead to any overall decline in religious participation, but rather will lead to increased members for organizations that actively adapt in ways that are responsive to contemporary needs. Concomitantly, those organizations failing to adapt would be expected to lose members.

The fact that evangelical churches are thriving makes sense within this framework. They have retained a distinctive lifestyle and "brand" identity by endorsing a rhetoric of family values in the face of rapid change, but they have adapted to changes in work and family well on a practical, every day level. This is evidenced by congregations that develop programs for dual-earner couples and by religious leaders who urge men to become more involved fathers and husbands, and to temper in practice the claims of male "headship" of the family that the official rhetoric endorses. Evangelicals have not adapted on issues of homosexual unions or by endorsing co-habitation (but these groups are a small portion of the religious market).

More generally, at the local level, this theoretical framework would suggest that congregations thrive when they read local demands correctly, providing programming that is responsive to the work and family situations of those in their own communities, while carving out some distinctive niche that differentiates them from the other congregations and non-religious organizations that might also be offering services for families and children. This has implications not only across but also within religious traditions. So, even though mainline Protestant denominations have declined overall, one would expect some local congregations within mainline traditions to be thriving, having adapted successfully.

Congregations adapting to work and family would signal an overall modernization of the religious landscape, but rational choice approaches do not see modernization as dangerous or damaging for religious organizations. From the rational choice perspective, Weber's (1998) fears about the "disenchantment" of modern life have not been born out. Modernization fosters the pluralism of unregulated religious markets and free religious choice, leading to vital religious belief and expression. Rational choice theorists explicitly reject the teleology of decline that underlies secularization theory, and argue that modernization has beneficial effects on rates of religious participation despite (in fact, because of) the weakening of official religious influence in government and other institutions.

In short, rational choice theories would tend to predict a great deal of change and adaptation among religious organizations in the face of changes in work and family, along with some shifting of who is religiously involved. It is irrational for religious organizations not to adapt, and the theory provides only a few explanations for any failure to adapt These mostly center around blocks to effective communication of market demand or effective decision-making in response to it. For example, a bureaucratic structure that would remove religious leaders from the immediacy of consumer demands in a hierarchical denomination might delay or forestall adaptation - and lead to decline.

In contrast to this picture of a highly adaptive market with rapid and responsive change, an institutional approach tends to emphasize stability and continuity of religious organizations. Religious institutions are bundles of ideas and practices that have come, over time, to be valued for their own sake. Current institutional arrangements guide and direct which social changes will be attended to, and foster standard ways of responding to change; all of this may or may not be "adaptive," from a market perspective.

An institutional perspective would begin by emphasizing the institutional filters that set limits on the changes that religious organizations and leaders are even likely to consider as relevant for their own organization. For example, where a rational choice perspective emphasizes the "practical egalitarianism" of evangelicals, an institutional approach would emphasize a core religious ideology that makes any serious consideration of women's structural equality with men illegitimate within these religious organizations. Evangelical feminists aside, there are few churches within evangelical (or holiness) traditions that accept women pastors, and none within the fundamentalist camp. Does this provide a limit on their organizational expansion in the face of increasing belief in women's fundamental equality in the United States? To take another example, some mainline Protestant leaders may want to adapt to changes in work and family but be hampered by an institutional ideology that de-values a focus on "the private" and worries about "needy people" sapping the church's energy for social justice ministries.

However, institutional theories do provide ways of thinking about when and how adaptation might occur. Theorists as diverse as Sewell (1992), Stryker (1994), and Friedland and Alford (1991) note that institutions change when they come into contact with other institutions that provide different frameworks for thinking about the world (what Sewell calls schema) and different rules of appropriate action. In this, institutional theory makes a place for agents who, while not as unconstrained as the typical rational actor of market approaches, nevertheless have considerable scope for creativity and for affecting the organizations and institutions of which they are a part. Religious leaders can try to change their organizations, or found new ones. Members of religious groups can encounter new ideas, whether from other religious organizations or other institutional arenas - like work, or education, or family - and can apply these to the religious organizations of which they are members, seeking change.

However, institutional theories do not rely as heavily on entrepreneurs in explaining innovations. Rather, changes tend to flow through standard institutional channels - for example, seminaries that train pastors in certain forms of ministry, ecumenical societies where pastors meet and discuss new programs, or networks of local churches who band together to cooperate in specific ministry areas. When faced with a decision, the institutional actor may not deploy the response likely to generate the greatest benefit for the smallest cost, but may rather deploy the response that is best known, most familiar, or most likely to please powerful outsiders.

Adaptation then, when it does occur, may not be an immediate and transparent response to members' or potential members' demands. And adaptation may not work; changes can be unresponsive to the needs of members and potential members. Of course, market theories can account for this too, but why this happens is left un-theorized, a "black box" of irrational behavior. Institutional theories provide ways to think about the social processes that may influence the "how" of adaptation, like the desire to maintain institutional routines or protect the legitimacy, authority, or interests of those in power.

Much adaptation, in this view, is undertaken to solve immediate problems, and there may be little attention to the long-term or unintended consequences of today's innovation. A conservative Baptist church board that votes to allow women to serve as elders one day may find itself pressured to re-think the issue of women pastors the next (Becker 1999). "Innovator" churches may find themselves asked to continue to innovate in ways they did not anticipate; and congregational leaders may resist all innovation, knowing about the "slippery slope" they may encounter if they begin.

It is important to remember that the agents of institutional theory are not the rational actors of market approaches. Institutional theory leaves open the goals of social action in the way that stricter approaches to rational choice theory do not. Rational choice approaches assume that agents pursue wealth or status, and that they engage in an instrumental calculation of the costs and benefits of their actions (means-ends rationality). An institutional approach leaves open the possibility that agents might choose to engage in value rational action, or action to pursue ends defined as right or good for their own sake; such goods are incommensurate, and may be pursued regardless of cost.

Because of this, institutional theory does not frame a decision not to adapt as a failure of rationality. Thus, it provides a way to account for a decision by a religious leader to risk a loss of membership or money - a decline in market share - in order to preserve something deemed more valuable - for example, doctrinal orthodoxy or right practice. In this, an institutional approach does not lead to the assumption of continued adaptation in the face of changes in work and family. And it is more open to the idea that some changes may lead to an "accommodation" of modernity that can lead to overall secularization and the decline of religious authority.

I will argue that an institutional approach provides a better framework within which to understand the dynamics of change in the religion-family nexus since the 1950s. This argument is based upon two kinds of supports. The first is theoretical, and the second is substantive.

In terms of theoretical scope and power, market approaches suffer in comparison to institutional approaches. Market theories have led to useful insights about how agents work to achieve their own interests, and hence to bring about change, and they have provided a valuable push in the sociology of religion to re-center the role of individuals and their choices in understanding religious affiliation and commitment. But they cannot explain without borrowing heavily from other theoretical approaches, including institutional theory, where preferences come from, or why actors may choose to maximize things other than prestige and wealth. Market approaches do not adequately account for the role of institutional filters in selecting the social changes to which agents respond, or the institutional mechanisms that lead to the deployment of standard kinds of solutions to problems of adaptation, regardless of whether they are effective from a market standpoint.

Institutional theories can take into account most of the same issues that market theories address. They also address important issues upon which market theories are silent. Institutional theories make a place for agents pursuing their interests, while explicitly theorizing the social and cultural mechanisms that lead to a non-random (institutionalized) pattern of preferences and interests in a society at a given time. They make a place for value-rational as well as instrumental-rational action, and they provide a framework that balances a focus on adaptability and change with an equal focus on social constraint.

Substantively, I will argue that an institutional framework is the best way to explain the pattern of findings in the four upstate New York communities that I studied in the course of this research. Most of the congregations in these communities are still using the "standard package"of family ministry developed in the 1950s as the basis for their weekly ministry. Most innovations take the form of one or two minor additions to this standard package. For example, single mothers will be individually encouraged to attend the "regular" mothers' group, or newly divorced members will be invited to the singles' outings. Rather than the rapid and responsive adaptation predicted by market theories, there are more gradual and incremental forms of adaptation based on previously institutionalized organizational routines and programming packages.

Also, the overall family orientation in local congregations - the combination of family rhetoric, programming, and ministry practices - varies according to the congregation's religious tradition. In addition to an evangelical family orientation, there are distinct liberal Protestant and Catholic patterns of response to change in work and family. Daycare centers are a liberal Protestant innovation; Catholics have taken the lead in ministering to blended families and the children of divorced parents; evangelical congregations do the most to be supportive of single mothers. This supports the idea that innovations do not occur in response to local market demand but are generated within and flow along institutional structures like seminary training programs.

Finally, some adaptations are clearly undertaken and pursued regardless of market effectiveness. Some liberal Protestant pastors reject the term "family ministry" entirely and talk about "community," because they feel that the traditional congregational focus on the nuclear family can too easily be interpreted as anti-feminist and anti-homosexual, and because it leaves out the increasing numbers of single adults and single parent families. Most have found this a somewhat unpopular stand to take; it is a position of principle, not a marketing strategy, and it has cost some congregations members.

This is not to say that market approaches are not useful in understanding some aspects of religious institutional change. For example, there are a few "innovator" congregations in these four communities which have deployed multiple new forms of family ministry across issue areas. They are all very large congregations with a thriving, younger membership. This could be interpreted to mean that adaptation can bring about organizational success and increased market share, which rational choice theory would predict. In addition, quite a few pastors are employing a discourse of competition and marketing in describing their attempts to adapt to changes in work and family, especially in talking about innovations in the time and timing of ministries and programs, something that is not an ideologically sensitive issue. Over 45% of pastors report that their congregation has experimented with the timing of the main worship service or other ministries in an attempt to draw in more members or to be more responsive to current members' needs. However, the overall picture in these four communities is more compatible with an institutional account of processes of change, rather than the fluid, dynamic, an innovative market that rational choice accounts would predict.

The theories above speak mainly to how change comes about, but they have different implications for the end result of religious institutional change. Market theories predict a great deal of adaptation and even modernization resulting from changes in work and family, but they reject the idea that modernization leads automatically to secularization. In support of this they cite the high and stable rates of religious participation in the United States over the course of this century. Institutional theories predict two kinds of responses. Some congregations would be expected to "resist" modernization by refusing to adapt, while others would accommodate modern work/family arrangements and would be judged as more secular, regardless of market success. In any case, institutional theories treat the question of secularization and the question of market success as two separate issues, and one cannot be induced in any straightforward way from the other.

There are many different versions of secularization theory, but they all have in common the idea that modernization undermines traditional religious authority. This decline of authority can exist at all levels of analysis. Secularization might mean that political leaders and institutions are less influenced by religious groups and leaders, that religiously-based prohibitions on the sale of alcohol on Sundays are lifted, or that individuals increasingly live their lives and make decisions without following the advice of religious leaders or the prescriptions of religious doctrines. It might also mean that religious leaders change what they teach over time, bringing faith traditions into line with modern or progressive attitudes and beliefs and eschewing a more orthodox view of correct religious doctrine or practice.

Some have argued, however, that taking a particular historical position as "orthodox" and interpreting any change from this position to be a "decline of religious authority" imposes a teleology on historical changes that may, in fact, not be unidirectional. In other words, not all religious change automatically signals a decline in religious authority. Theoretically, there can be changes that indicate a re-sacralization of social life. There can also be changes in which any kind of directionality (more or less secularization) is more difficult to discern.

From this perspective, those who argue that any particular change equals "modernization" and therefore "secularization" need to be able to defend their argument by demonstrating that the content of change amounts to a more "modern" outcome by some a priori criterion with which most scholars would agree. For example, conflicts over ordination of women and homosexuals into the priesthood seem clearly to be struggles over accepting modern standards of judgment over traditional ones. But a congregation changing its ministry to accommodate two-career couples is only modernization - or secularization - if a nineteenth-century understanding of appropriate gender roles is designated as "traditional" or "orthodox." But since the male-breadwinner family is itself a modern artifact of the industrial revolution, this chain of reasoning seems much less convincing. So, congregations that adapt to a dual-earner lifestyle may be more contemporary but they are not necessarily more modern or more secular.

Drawing on Bourdieu (1977), Kurtz (1986) offers an alternative framework with which to understand changes in religious authority that avoids the teleology of decline embedded within secularization theory. Rather than positing religious orthodoxy as a "pure" or "traditional" state from which all change must be understood as secularization, religious orthodoxy is understood to be in all situations historically constructed and defined in relation to a heterodox alternative. In the political struggle between those defending an orthodox position on any given issue and their heterodox challengers, the orthodox is itself changed, incorporating some of the positions taken by its heterodox critics. Orthodoxy, then, is best understood as an official interpretation of religious truth that is always evolving and changing as particular actors seek to influence it according to their own socially-located experiences and understandings of the world.

The main advantage of an "evolving orthodoxy" view is that it provides a theoretical language for talking about religious institutional change as a continuous historical process without imposing a teleology of decline and loss. The content or direction of change - what it all adds up to - is left open, and the question of secularization then becomes an empirical one.

So, if religious institutions have adapted to changes in work and family, an evolving orthodoxy approach does not assume such adaptation signals accommodation to modernity or the secularization of religious authority. Rather, one would have to specify the results of the changes in order to understand if they resulted in religious institutions that exercised less authority in some area of life than was previously the case.

I will argue that an evolving orthodoxy framework is a better explanation for the religious institutional changes that I observed in the four communities in which I conducted research for the Religion and Family Project. In the vast majority of congregations an intact nuclear family is still not only the ideal but also the unit around which congregational ministries and practices are organized. There is little formal or informal acknowledgement of gay and lesbian members in these congregations, all of it located among a few liberal Protestant churches. Ministry to single parents exists, but is mostly being carried out through more traditional programs like mothers' groups. In short, there is no decisive movement to a more modern understanding of 'the family' in local congregational life among most congregations. And many changes - having baby-sitting during Sunday worship, holding parenting classes, providing referrals for substance abuse or domestic violence counseling -- simply do not seem to reflect a "modernization/traditionalism" dynamic.

On the other hand, it does seem to be the case that increasing involvement in paid work is effecting some changes that could be understood as secularization. Those working more than 51 hours a week are significantly less likely to be church attenders in my sample than those working fewer hours. Women's religious involvement is still predicted by family formation, especially marriage and the birth of younger children, but married women who work full time are less religiously involved than married women who work fewer hours, or those who are homemakers. Women's own attitudes and beliefs are more strongly predictive of their religious involvement than are men's. Specifically, women in these communities are higher than men on measures of religious individualism and privatism, and these attitudes influence their religious involvement more than they influence men's involvement.

Taken together, there is some support for the interpretation that women's increased labor-force participation has led them to a more problematic relationship with religious institutions, and that this is particularly true for married women (cf. Hertel 1995). And for both men and women, those scoring high in religious individualism and privatism have reduced religious involvement in these communities. Both of these developments might be a sign of secularization in the sense of modern, individualistic activities and attitudes undermining religious participation.

This introduction makes clear that there are many more questions than answers in our understanding of the relationship between religious institutions and families after a period of social change. In the next chapter, I will discuss in more detail the design of the study, the four communities chosen, and the strategies of collecting data for the Religion and Family project, and explain why this study provides a good basis for beginning to develop some answers to the questions posed above. Chapter 2 also discusses some issues of measurement and method. How do we measure religious involvement and what are the consequences of using different kinds of religious involvement measures? How can we measure religious individualism, a concept that has received much attention in the literature on religious institutional change?

The rest of the manuscript is divided into three parts. Section II profiles the religious involvement of individuals in Tompkins County, Seneca County, Liverpool, and Northside. In Chapter 3, I discuss how individuals' family context and position in the life-course are related to religious involvement. This chapter profiles how traditional family formation events - marriage, the birth of children - are related to involvement, but also looks at the effects of family disruption and alternative life-course pathways. Chapter 4 analyzes how commitment to and involvement in work - paid labor - is related to religious commitment and involvement for individuals in these four communities. This chapter provides some support for the "time bind" thesis, but also argues that, for some, religious involvement leads to "scaling back" at work to spend more time with family. Chapter 5 takes up the question of the gendering of religious involvement after a period of rapid change in work and family, and explores the different determinants of and meaning of religious involvement for men and women.

Section III outlines the religious-institutional response to changes in work and family.

Chapter 6 looks at congregational ideology and discourse. What ideas about women's roles, about children, about gay and lesbian unions, are promoted within congregations? What ideals of the good family are promoted? Chapter 7 looks at the range of family programming in congregations. If ideals can be made explicit in rhetoric, they are also embodied in the structure of programming and in organizational routines. What do congregation's programs and ways of doing things communicate about the kinds of people who are welcomed and valued in congregational life? Chapter 8 sums up the overall styles of innovation across religious tradition and looks at the pathways to innovation in family ministry and family rhetoric.

Section IV reassesses the theoretical explanations discussed briefly above. Chapter 9 develops the evidence for an understanding of religious institutions as engaged in an "evolving orthodoxy" in their overall family orientation. Chapter 10 assesses the influence of market and institutional processes of change, and argues that overall it is more fruitful to think about religious institutions as operating in an institutional field, rather than in a market. Chapter 11 examines the public implications of the how congregations have adapted to changes in work and family, and of the kinds of rhetorics they provide for thinking about appropriate and moral family lifestyles in contemporary society.

+ U.S. Bureau of the Census, http://www.census.gov/population/socdemo/hh-fam/htabHH-1.txt,

accessed 12/21/99, and http://www.census.gov/population/socdemo/hh-fam/htabFM-1.txt, accessed

01/23/00.

ƒ Defined as two or more related persons living together.

* Hayghe, Howard V. and Susan M. Bianchi, "Married Mothers' Work Patterns," Monthly Labor

Review, June 1994:24-30.