xiv Preface and Acknowledgments colleagues and students at the John F. Kennedy School of Government, Harvard University, and the Department of Political Science and the Institute for Social Research at the University of Michigan. Ann Arbor, Michigan Cambridge, Massachusetts May 2002 PART I THE CAUSES OF THE RISING TIDE I Introduction Explaining the Rising Tide of Gender Equality During the late twentieth century, the issue of gender equality once again became a major issue on the global agenda. The UN Decade for Women, which ended in 1985, initiated the integration of women into development, triggering the formation of thousands of women's organizations and networking them across the world.1 The trend accelerated during the following decade. In 1993, the Vienna World Conference proclaimed that women's rights were human rights; in 1994, the Cairo International Conference on Population and Development placed women's empowerment and health at the center of sustainable development programs. Two years later, the Beijing Fourth World Conference on Women adopted a platform seeking to promote and protect the full enjoyment of human rights and fundamental freedoms for all women.2 Although there has been substantial progress toward gender equality in much of the world, great disparities persist, as systematic indicators demonstrate.3 In many places, most women's lives remain wretched. Afghanistan was among the most oppressive regimes, with women and girls living under an extreme version of Islamic law introduced by the Taliban. They were denied education, barred from the workplace, and unable to venture out in public without a male companion and the full head-to-toe covering of the burqa. They suffered from limited access to health care, including laws forbidding treatment of women by male doctors, and pervasive threats of domestic and state-legitimated violence.4 Few regimes are so draconian, but women in many societies face endemic 3 4 I. The Causes of the Rising Tide and substantial gender gaps in the division of household responsibilities, limited access to educational opportunities and economic resources, as well as legal and political barriers to positions of political power. Indicators of well-being ranging from literacy and longevity to labor force participation, poverty rates, and child mortality and schooling all reveal persistent disparities between women and men.5 Some societies have experienced a mixture of progress and regression, as new entrepreneurial opportunities arose for women following market liberalization in post-Communist Europe, along with weakened social safety nets for poorer families.6 By contrast, other countries have achieved major gains in legal, economic, and political gender equality that are probably irreversible. Sweden exemplifies a society where women experience the highest level of parliamentary representation of any nation in the world, along with gender parity in secondary schooling and paid employment and extensive parental rights and childcare facilities.7 Although such contrasts in women's lives around the globe are well established, the reasons for them are not. What explains the disparities between the leaders and laggards in gender equality? Economic Growth and Human Development One approach common the 1960s and early 1970s emphasized economic growth as the most effective strategy for achieving human development and improvements in the living conditions and status of women. After World War II, optimism abounded that the world could be rebuilt to end poverty, injustice, and ignorance, improving women's lives as an inevitable part of development. Walt Rostow's influential book The Stages of Economic Growth (i960) suggested that human progress was driven by a dialectic that could be accelerated.8 The end of colonial rule in many parts of Africa, Asia, and the Caribbean was seen as a major opportunity to promote prosperity and democracy in these societies. Greater affluence was expected to facilitate freedom from want and fear due to an expansion of health care and adequate nutrition, schools and housing, jobs and basic social protections, increasing the urban middle classes and laying the social foundations for the consolidation of democratic institutions and civic society. Growth was seen as the panacea that would lift all boats, and it was often Introduction 5 implicitly assumed that this included endemic problems of women's literacy and education, their poverty, low pay, and occupational segregation in the workforce, their care-giving responsibilities in the home and family, and their participation and representation in the political system. The hope that economic development will automatically benefit women in poorer societies continues to be voiced.9 At its most simple, this proposition is often taken for granted as self-evident. After all, in the examples we have cited, Sweden is one of the richest societies in the world, with a per capita income of $26,000 per year; the figure for Afghanistan is around $800. Do these countries' striking differences in gender equality simply reflect their differing degrees of development? But by the end of the twentieth century, the limitations of growth alone were clear. Numerous anomalies are obvious even to the casual observer. Kuwait, Saudi Arabia, and Quatar, for instance, are about as rich as Sweden in per capita GDP, but women in these societies cannot stand for office or even vote, and they have narrowly restricted rights and opportunities outside the home. It is illegal for women to drive in Saudi Arabia, and the Middle East and North Africa have the lowest rates of female labor force participation in the world.10 Conditions for women are more favorable in some poorer nations. In India, for example, although women's rights are also limited in many important ways, about 800,000 women serve in local government, with one-third of all local council seats reserved for them.11 Broader experience confirms that gender equality in elected office continues to lag behind in the transitional "Asian tiger" nations, as well as in many high-growth states in Latin America.11 Even in the most affluent societies around the world, such as the United States, France, and Japan, where women have made substantial gains in access to universities, company boardrooms, and the professions, there has been minimal progress for women in government - while in South Africa, by contrast, women comprise almost one-third of all parliamentarians, ranking this nation eleventh worldwide in the proportion of women in the lower house.13 It has become apparent that problems of gender equality are more complex and intractable than the early developmental theorists assumed.14 Growing affluence does tend to generate the expansion of literacy and schooling, the establishment of a social protection safety net, and the rise of white-collar jobs in the service sector, but this 6 I. The Causes of the Rising Tide process is not inevitable, nor does it necessarily automatically benefit women's lives. The Role of the State: Human Rights, Legal Reforms, and Political Institutions During the 1980s and 1990s, recognizing the limitations of economic strategies alone, the international women's movement and official bodies such as the United Nations and European Union turned their attention increasingly toward the role of the state in reinforcing or alleviating institutional barriers to women's progress, and toward the need to establish political, social, and economic rights in order to secure gender equality through legal reform and the courts.15 There was also a shift in the literature around this period, from focusing on the problems facing women's well-being toward emphasizing the active role of women's agency and voice in helping women to attain equal rights: to earn an independent income, to find employment outside the home, to have ownership rights, to become literate, to participate in community decision making.16 The independence and empowerment of women became understood as an integral part of the development process, so that women could articulate their own wants and needs. In many countries legal rights for women remain limited; a comprehensive review of legislation in over 100 countries by Humana found that in the early 1990s women still lacked many basic rights, such as the right to own land, to manage property, to conduct business, and even to travel without spousal consent.17 In much of sub-Saharan Africa, women have land rights through their husbands as long as the marriage endures, but lose this property when divorced or widowed. In Turkey, until a recent reform of the civil code, a wife needed her husband's consent to work outside the home; women were not entitled to sue for divorce, to claim alimony, or to retain their maiden names. In Egypt and Jordan, women need their husband's permission to travel. In Ireland, it remains illegal to have an abortion except in extremely limited circumstances (where the mother's life is in danger). In established democracies, women have had the legal franchise for many decades -since the 1920s in most Protestant countries, and since the 1950s in most Catholic ones. But in newer democracies, such as Namibia and South Africa, most women have only recently acquired voting rights. Introduction 7 And laws restricting women's rights to vote and to run for office persist in a handful of Middle Eastern countries, including Kuwait, Qatar, Saudi Arabia, Oman, and the United Arab Emirates.18 The United Nations has encouraged states to recognize women's rights, most importantly through the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW), adopted by the UN General Assembly in 1979 and subsequently signed by 165 nation-states. This convention emphasizes the importance of women's equal participation with men in public life. The women's movement in many nations has emphasized the need for equal opportunity and affirmative action strategies through reforming institutional barriers, removing structural biases, and altering the rules of the game to get women into positions of elected office. A particularly effective means to do this has been the use of quotas in the selection of female parliamentary candidates - which has recently been adopted in many Western European, Asian, and Latin American countries - and the parity program adopted in France.19 Policies designed to prevent sex discrimination, to secure equal pay, maternity, and reproductive rights, and to increase opportunities for women in the workforce and education have been adopted in many countries, and the role of the state is now widely understood to be central in actively consolidating and reinforcing gender equality.20 These strategies have secured concrete gains for women in many nations, particularly when government agencies or the courts have effectively implemented legal reforms and policy initiatives. Changing the "rules of the game" can have a dramatic impact on women's lives, accelerating progress and opening new opportunities. Yet at the same time there can be a substantial gap between the recognition of de jure formal rights and actual practice. Many governments have signed international conventions pledging themselves to support equal opportunity in political representation; and political leaders, official bodies, and administrative agencies have often declared themselves in favor of this principle, along with groups in civic society such as trade unions and parties. Yet in the world as a whole, women remain far from parity at the apex of power - as heads of state at the prime ministerial and presidential level, in the executive branch as ministers and as senior public officials, and in national parliaments.2-1 In the same way, CEDAW recognizes the importance of equality in the paid labor force. 8 I. The Causes of the Rising Tide Yet although many governments have signed on to this principle, in practice women are disproportionately likely to have low-wage jobs because of persistent occupational segregation and wage discrimination by sex, as well as lack of child care for working mothers; and in most countries women in management and corporate boardrooms continue to encounter a glass ceiling.22 Even in liberal countries such as Sweden and Norway, segregation in jobs typically held by women and men remains common. Statutory reform and formal recognition of women's rights in international bodies are symbolic gains, an important advance in itself, but they are seldom sufficient to effect substantial social change if the capacity or the political will to implement these reforms remains weak. Cultural Barriers Economic growth and legal-institutional reforms are both important in any long-term comprehensive strategy to promote gender equality. But in addition, as this book will demonstrate, culture matters, and indeed it matters a lot.23 Perceptions of the appropriate division of roles in the home and family, paid employment, and the political sphere are shaped by the predominant culture - the social norms, beliefs, and values existing in any society, which in turn rest on levels of societal modernization and religious traditions. 'Gender' refers to the socially constructed roles and learned behavior of women and men associated with the biological characteristics of females and males.24 In many societies, rigid gender roles determine the rights, resources, and powers of women and men, notably the division of labor in the home and workplace. In others, men's and women's roles are more interchangeable, and innate biological differences lead to fewer social expectations. Where a culture of gender equality predominates, it provides a climate where de jure legal rights are more likely to be translated into de facto rights in practice; where institutional reforms are implemented in the workplace and public sphere; where women embrace expanded opportunities to attain literacy, education, and employment; and where the traditional roles of women and men are transformed within the household and family. Moreover, the critical importance of culture is that women as well as men adopt the predominant attitudes, values, and beliefs about the appropriate division of sex roles within any society. Introduction 9 Where traditional values prevail, women are not only limited by society in terms of the opportunities they seek, but also choose to limit themselves. Cultural change is not sufficient by itself for gender equality - a limitation not always sufficiently recognized by the consciousness-raising individualistic focus of the women's movement of the 1960s. But we argue that cultural change is a necessary condition for gender equality: women first need to change themselves before they can hope to change society. In turn, cultural change lays the basis for the mass mobilization of women's movements and broad support for public policies that reinforce, consolidate, and accelerate the process of gender equality. At one level, there is nothing particularly new or startling about this claim. A mainstream tradition in sociology, anthropology, history, and social psychology has long theorized that there are great cross-cultural differences in beliefs about gender roles among societies around the globe, even among societies at similar levels of socioeconomic development - such as Sweden, Britain, and the United States, on one hand, and India, the Philippines, and Indonesia, on the other.25 Feminist movements in many countries have long emphasized cultural differences in family and sex roles, and the critical importance of changing traditional patriarchal norms for transforming relationships between the sexes.26 Most support for this thesis has come from qualitative evidence, often based on personal interviews, participant observation, and case studies. Comparative analysis of aggregate indicators has also revealed the substantial contrasts between the lives and roles of women and men worldwide. Nevertheless, systematic survey evidence monitoring cultural attitudes toward gender equality across many societies remains scattered and inconclusive, with most studies limited to a handful of affluent postindustrial societies and established democracies in Western Europe and North America.27 While it is widely assumed that culture matters, it remains unclear how much it matters as compared to levels of societal development and legal-institutional structures; and we know even less about how these factors interact in the long-term process of value change. This book will demonstrate that cultural traditions are remarkably enduring in shaping men's and women's worldviews; nevertheless, glacial shifts are taking place that move systematically away from traditional values and toward more egalitarian sex roles. This shift is intimately related to the processes of societal modernization lO I. The Causes of the Rising Tide and to generational replacement. Moreover, we will demonstrate that culture matters: where there are more egalitarian attitudes, these are systematically related to the actual conditions of women's and men's lives. We acknowledge that this is not a simple one-way direction of causality; rather, it is an interactive process, because changes in our lives affect our underlying attitudes and values. But we also demonstrate that cultural change is not an ad hoc and erratic process; rather, patterns of human development and societal modernization underpin attitudinal shifts. The broad direction of value change is predictable, although the pace is conditioned by the cultural legacy and institutional structure of any given society, as exemplified by the role of an Islamic heritage in the Middle East, the legacy of Communism in Central Europe, and the egalitarian tradition in Scandinavia. To develop these arguments, this book examines evidence of a rising tide of support for gender equality in over seventy societies around the world. It then explores the causes of this cultural shift and its consequences for women's political power, including their civic engagement, support for the women's movement, and political representation. This introductory chapter first develops the core theoretical argument and outlines the research design, providing details about (i) the four waves of the World Values Survey / European Values Survey carried out from 1981 to 2001, (2) the comparative framework and societal classification used here, and (3) the time period used for trend analysis. The final section of the chapter outlines the book and summarizes the contents of subsequent chapters. Societal Modernization and Cultural Change The revised version of modernization theory developed in this book hypothesizes that human development brings changed cultural attitudes toward gender equality in virtually any society that experiences the various forms of modernization linked with economic development. Modernization brings systematic, predictable changes in gender roles. The impact of modernization operates in two key phases: 1. Industrialization brings women into the paid workforce and dramatically reduces fertility rates. Women attain literacy and greater educational opportunities. Women are enfranchised Introduction u and begin to participate in representative government, but still have far less power than men. 2. The postindustrial phase brings a shift toward greater gender equality as women rise in management and the professions and gain political influence within elected and appointed bodies. Over half of the world has not yet entered this phase; only the more advanced industrial societies are currently moving on this trajectory. These two phases correspond to two major dimensions of cross-cultural variation that will be described in more detail in the final chapter: (1) a transition from traditional to secular-rational values, and (2) a transition from survival to self-expression values. The decline of the traditional family is linked with the first dimension. The rise of gender equality is linked with the second. Cultural shifts in modern societies are not sufficient by themselves to guarantee women equality across all major dimensions of life; nevertheless, through underpinning structural reforms and women's rights, they greatly facilitate this process. Modernization theories suggest that economic, cultural, and political changes go together in coherent ways, so that industrialization brings broadly similar trajectories even if situation-specific factors make it impossible to predict exactly what will happen in a given society. Certain changes become increasingly likely to occur, but the changes are probabilistic, not deterministic. Modernization theories originated in the work of Karl Marx, Max Weber, and Emile Durkheim. These ideas were revived and popularized during the late 1950s and early 1960s by Seymour Martin Lipset, Daniel Lerner, Walt Rostow, and Karl Deutsch.28 These writers argued that the shift from agrarian agriculture towards industrial production leads to growing prosperity, higher levels of education, and urbanization, which in turn lay the social foundations for democratic participation in the political system. Traditional societies are characterized by subsistence economies largely based on farming, fishing, extraction, and unskilled work, with low levels of literacy and education, predominately agrarian populations, minimum standards of living, and restricted social and geographic mobility. Citizens in agrarian societies are strongly rooted to local communities through ties of "blood and belonging," including those of kinship, family, ethnicity, and religion, as well as through 12 I. The Causes of the Rising Tide long-standing cultural bonds. The shift from traditional agrarian society toward industrialized society involves the move from agricultural production to manufacturing, from farms to factories, from peasants to workers. Social trends accompanying these developments, as shown in Table i.i, include migration to metropolitan conurbations, the rise of the working class and urban bourgeoisie, rising living standards, the separation of church and state, increasing penetration of the mass media, the growth of Weberian bureaucratization and rational-legal authority in the state, the foundations of the early welfare state, and the spread of primary schooling. This phase occurred in the Industrial Revolution in Britain during the mid to late eighteenth century and spread throughout the Western world during the nineteenth and early twentieth centuries. The early developmental theorists emphasized a range of social trends that commonly accompany the process of industrialization, including changes in traditional sex roles, the family, and marriage. During the early 1970s, Daniel Bell popularized the view that after a certain period of industrialization, a further distinct stage of development could be distinguished, as a nonlinear process, in the rise of postindustrial societies.29 For Bell, the critical tipping point was reached when the majority of the workforce moved from manufacturing into the service sector, working as lawyers, bankers, financial analysts, technologists, scientists, and professionals employed in the knowledge industries. The now-familiar social and economic shifts characterizing postindustrial societies are listed in Table i.i They include the rise of a highly educated, skilled, and specialized workforce; population shifts from urban to suburban neighborhoods and greater geographic mobility, including immigration across national borders; rising living standards and growing leisure time; rapid scientific and technological innovation; the expansion and fragmentation of mass media channels, technologies, and markets; the growth of multilay-ered governance, with power shifting away from the nation-state toward global and local levels; market liberalization and the expansion of nonprofit social protection schemes; the erosion of the traditional nuclear family; and growing equality of sex roles within the home, family, and workforce. There is a broad consensus that certain socioeconomic developments have been sweeping across many societies, although alternative Introduction 13 table i.i. Typology of stages of societal modernization From Agrarian to Industrial Societies From Industrial to Postindustrial Societies Population Human capital Workforce Social status Living conditions Science and religion The population shift from agrarian villages to metropolitan conurbations. Growing levels of education, literacy, and numeracy and the spread of basic schooling. The shift from extraction and agriculture toward manufacturing and processing. The rise of the working class and the urban bourgeoisie, and the decline of peasant society and traditional landed interests. Growing standards of living, rising longevity and expanding leisure time. The industrial revolution in manufacturing production. Growing division of church and state. The diversification of religious sects and denominations. The diffusion from urban areas to suburban neighborhoods. Greater social geographic mobility, including immigration across national borders, generating the rise of more multicultural societies. Rising levels of education, especially at secondary and university levels, generating increased levels of human capital and cognitive skills. The rise of the professional and managerial occupations in the private and public sectors and greater occupational specialization. The move from ascribed occupational and social roles assigned at birth toward achieved status derived from formal educational qualifications and careers. Economic growth fueling an expanded middle class, rising living standards, improved longevity and health, and growing leisure time. Rapid technological and scientific innovation. The process of secularization weakening religious authority. (continued) 14 I. The Causes of the Rising Tide TABLE i.i (continued) From Agrarian to Industrial Societies From Industrial to Postindustrial Societies Mass media Government Social protection Family structures Sex roles Cultural values The wider availability of mass-circulation newspapers and periodicals and, during the twentieth century, access to electronic mass media. The expansion of the franchise, the growth of Weberian bureaucratization and reliance on legal-rational authority in government. The development of the early foundations of the welfare state and the elements of social protection for illness, unemployment, and old age. The shift from extended to nuclear families, the gradual reduction in the fertility rate. The entry of more women into the paid workforce. Material security, traditional authority, and communal obligations. The shift in the mass media from mass broadcasting toward more specialized narrowcasting, with the fragmentation of media outlets across markets and technologies. The growth of multilayered governance at the global and local levels, as well as the expansion in the nonprofit sector. Market liberalization and the contraction of the state, displacing social protection increasingly to the nonprofit and private sectors. The erosion of the nuclear family, the growth of nontraditional households, and changing patterns of marriage and divorce. Growing equality of sex roles in the division of labor within the home, family, and workplace, and the rise of women (especially married women) in the paid labor force. Quality of life issues, self-expression, individualism, and postmaterialism. Introduction 15 interpretations dispute their exact nature and periodization and the appropriate weight to be given to different components. There remains considerable controversy, however, concerning the consequences of these changes, and in particular concerning the probable impact of the modernization process on gender equality. Why would we expect these changes in socioeconomic conditions to go hand in hand with cultural shifts? In a series of works, Inglehart has demonstrated how the evolution from agrarian to industrial to postindustrial societies brings about two coherent, predictable, and interrelated dimensions of change: (1) socioeconomic changes in the process of production, as Bell claimed; and (2) a transformation in societal cultures, including rising emphasis in postindustrial societies on the pursuit of quality-of-life values rather than material concerns.30 We see economic, political, and cultural changes as evolving together in coherent trajectories, without claiming, as Marx did, that the changes in the processes of economic production drive the superstructure of value change, or that, conversely, cultural processes such as the rise of Protestantism cause the socioeconomic developments, as Weber argued. We view these causal processes as reciprocal. People living near the subsistence level tend to be primarily concerned with the basic struggle for survival when facing the unpredictable risks of disease, illiteracy, malnutrition, infant mortality, ethnic conflict and civil war, unsafe drinking water, and the spread of AIDS/HIV. Women and children are among the most vulnerable populations in these societies, not only because they are high-risk populations but also because they are usually dependent on a male breadwinner. Of the world's 6 billion people, the World Bank estimates that 1.2 billion live on less than $1 a day.31 Global poverty fell substantially during the 1990s, mainly driven by high economic growth in some larger nations such as China and India, but extreme poverty in sub-Saharan Africa worsened. Levels of infant mortality have been reduced around the globe, but the problem remains substantial; in 1998, there were 105 deaths per 1,000 live births in societies with low human development.32 Basic problems of survival are starkly illustrated by average life expectancy; in agrarian societies, on average people can expect to live for fifty-nine years, twenty years less than in postindustrial societies. Opportunities for social and geographic mobility are i6 I. The Causes of the Rising Tide limited by minimal literacy and schooling, which is especially common for girls and women. In this context, Inglehart argues, poorer people in low-income societies tend to give top priority to meeting subsistence and survival needs. In richer countries, public policies provide subsistence-level incomes, adequate housing, and effective health care. In preindustrial societies, without state services to cushion unpredictable blows, most people's lives are highly vulnerable to such risks as unemployment, ill health, poor crops, and violent crime, as well as to disasters such as floods, earthquakes, and famine. Societies whose people live with high levels of insecurity tend to develop cultures mistrustful of rapid change, emphasizing the values of traditional authority and strong leadership, inherited social status, and communal ties and obligations, backed up by social sanctions and norms derived from religious authorities. In these societies, the traditional two-parent family, with its division of sex roles between male breadwinner and female caregivers, is crucial for the survival of children, and therefore of society. Social norms buttress traditional family values and patriarchal norms of male dominance, strongly discouraging divorce, abortion, and homosexuality and instilling negative attitudes toward an independent economic role for women outside of the household. The legal structure involving property, marital, and citizenship rights for women reflects these traditional norms. Preindustrial societies emphasize childbearing and child rearing as the central goal of any woman, her most important function in life, and her greatest source of satisfaction and status. Given their very high rates of infant and child mortality, high rates of reproduction are emphasized in preindustrial societies; large extended families provide a source of subsistence and protection for the parents in old age, as well as a means for the transmission of land and property. In peasant societies, women usually work within the home, primarily in the production and preparation of food, and in child care. Even in these societies, there may be conflicts between the social norms shaping the appropriate division of sex roles and the actual life experiences of women and men, particularly if the male loses the capacity to act as the major breadwinner of the household, or if the woman heads a single-parent family.33 The rise of capitalism and the Industrial Revolution brought challenges to traditional values and a worldview that encouraged achieved Introduction 17 rather than ascribed status, individualism rather than community, innovation instead of continuity with tradition, and increasingly secular rather than religious social beliefs. The traditional roles of women -taking prime responsibility for care of children and the elderly -continued, but during the nineteenth and early twentieth centuries more and more women in industrial societies entered the paid labor force, mainly in factories and white-collar clerical and retail jobs, and attained greater legal rights to own property, to divorce, and to vote. Fertility rates and the size of the average family fell, reflecting the availability of safe contraception, improvements in health care that reduced the risks of infant mortality, and the fact that large extended families were no longer so crucial for protection in old age. People's lives became less vulnerable to sudden disaster, as savings and insurance schemes were developed to hedge against economic risks. The rise of local cooperatives, unions, building societies, and savings and loan schemes, and the development of social protection by philanthropic voluntary organizations and the state, helped to cover the worst problems of sickness, ill health, unemployment, and old age. In the period following World War II, postindustrial societies developed unprecedented levels of prosperity and economic security, with rising standards of living fuelled by steady economic growth, despite occasional cyclical downturns. Governments in these societies expanded the role of the welfare state to provide greater social protection for the worst-off citizens; more recently, health care, pensions, and care of the elderly have been contracted out to the nonprofit and private sectors, under state regulation. Under conditions of greater existential security, Inglehart theorizes, public concern about the material issues of unemployment, health care, and housing no longer necessarily takes top priority. Instead, in advanced industrial societies the public has given increasingly high priority to quality-of-life issues, individual autonomy and self-expression, the need for environmental protection, and direct participation in political decision making through petitions, protests, and demonstrations. Cultural shifts have transformed not only political life but also personal life, and nowhere more so than in the erosion of the traditional two-parent nuclear family; in liberalizing patterns of sexual behavior, marriage, and divorce; and in the wider acceptance among both women and men of greater gender equality in the home, the workforce, and the public sphere. Women are less i8 1. The Causes of the Rising Tide restricted to attaining status and fulfillment through the traditional route of family, marriage, and children, as alternative opportunities for self-expression and financial autonomy have become available. These changing norms have given rise to political demands, fuelling support for the second-wave feminist movement, and legal reforms associated with securing equal opportunities and women's rights. In short, the rising tide of support for gender equality in postindustrial societies is part of a broad and coherent cultural shift that is transforming economically developed societies. Although the broad outlines of this shift are predictable, not every society responds to these developments in the same way: as we will demonstrate, traditional cultural heritages help to shape contemporary social change. A society's values and religious beliefs, its institutions and leaders, and the structure of the state all help to shape this process in ways that differ from one society to another. Moreover, even in rich societies, some groups fall through the social safety net, producing disparities between rich and poor. If this theory is correct, and cultural shifts are coherent and predictable, then five specific propositions follow - each of which can (and will) be tested in this study. i. Cross-national comparisons: First, if coherent cultural patterns tend to be associated with specific levels of socioeconomic development, then postmodern values of gender equality will be most widespread in the most affluent and secure societies; conversely, the publics of poorer preindustrial societies will systematically be most likely to emphasize traditional gender roles. The fact that we have data from more than seventy countries, covering the full range of variance from low-income societies to affluent postindustrial societies, will make it possible to test this hypothesis in a more conclusive fashion than has ever before been possible. •>. Sectoral comparisons: Within any given society, postmodern values of gender equality will be most evident among the most secure, that is, the wealthier, better-educated sectors of the public. The less secure strata will prove more traditional in their attitudes toward women. 3. Gender comparisons: Women and men are expected to differ in their values and attitudes toward gender equality, with women proving more supportive of gender equality, especially Intra 19 in postindustrial societies. In traditional societies, both men and women often accept substantial gender inequalities. But societal modernization is transforming everyone's life experience, especially women's, reducing their vulnerability, generating greater financial autonomy, expanding literacy and educational opportunities, and strengthening the social safety net, especially for maternal and child care, reproductive control, and provision for the elderly. 4. Generational comparisons: In societies that have experienced sustained periods of economic growth and increasing physical security (such as Germany, the United States, and Japan) or very rapid economic growth (such as South Korea and Taiwan), we expect to find substantial differences in the values held by older and younger generations. The young should be more egalitarian in their attitudes toward sex roles, while the older cohorts should believe in more traditional roles for women and men; and the generation gap should be particularly large for women. We predict that the younger generation of women will hold the most egalitarian gender values within a given society, while the older generation of women will prove the most traditional. But societal values do not change overnight; instead, there is a substantial time lag, because adults tend to retain the norms, values, and beliefs that were instilled during their preadult years. Since, according to our hypotheses, these generational differences are linked to economic growth, we do not expect to find equally large generational differences in societies that have not experienced major increases in real GNP per capita over the last several decades (for example, Nigeria and Zimbabwe). 5. Religious legacies: Finally, we anticipate that religious legacies will leave a strong imprint on contemporary values. In particular, controlling for a society's level of GNP per capita and the structure of the workforce, we expect that the publics of Islamic societies will be less supportive of gender equality than the publics of other societies. In considering patterns of trends over time in changing values with respect to gender equality, the "convergence model" suggests that the changes in men and women's lives in the home, the workforce, and 20 I. The Causes of the Rising Tide the political sphere occurred first in postindustrial societies, driven by structural shifts in the workforce and educational opportunities, generating the women's movement of the 1960s and 1970s, but that in the long term laggard societies will gradually catch up as the culture shift ripples around the globe. The process of globalization has accelerated the recognition of women's rights, as well as the process of democratization. In the short term, however, value change has widened the cultural differences between postindustrial and agrarian societies, including the differences between women living in these types of societies. The Comparative Framework In order to examine the evidence for these predictions, this study follows Prezeworski and Teune's "most different systems" research design, seeking to maximize contrasts among a wide range of societies and thereby to distinguish systematic clusters of characteristics associated with different dimensions of gender equality.34 Some important trade-offs are involved in this approach, notably the loss of contextual depth that can come from focusing on one nation or studying a few similar countries. But the strategy of carrying out broad comparisons has major advantages. Most importantly, it allows us to examine whether, as theories of societal modernization claim, basic values seem to shift along with the shift from traditional agrarian societies, with largely illiterate and poor populations, through industrial economies based on manufacturing and a growing urban working class, to postindustrial economies based on a large service-sector middle class. And since it allows us to compare societies with sharply differing religious legacies, political systems, and democratic traditions, it enables us to analyze the role of these other, conceivably very important, factors. Human development is also a complicated, multifaceted process of social transformation, including changes in the economy, with the shift from agricultural production and extraction to industrial production and the rise of the service sector; changes in society, with the growth of education, affluence, and leisure, life expectancy and health, urbanization and suburbanization, the spread of the mass media, and changes in family structures and community social networks; and changes in politics, with the process of democratization. Not all of these developments necessarily go hand in hand in advancing the position of women Introduction 21 in every society. The early stages of industriah^uon, tor example, may expand literacy and educational opportunities for women, and yet may simultaneously weaken the informal extended family support networks available in agrarian communities. Structural adjustments in developing countries may produce efficiency gains in the longer term, but may also disproportionately hurt women's interests.35 The democratic transition in the post-Communist world opened up new opportunities for gains in political rights and civic liberties, but the abandonment of gender quotas for elected office simultaneously sharply reduced the number of women in parliament.36 In affluent nations such as the United States, Japan, and France, women have advanced in management and the professions further and faster than in legislatures and cabinets. One difficulty is that the abstract concept of "societal modernization" encompasses many complex dimensions of social change, including cross-cutting developments, some of which, like the growth of white-collar occupations and education, expand opportunities for women, while others, such as privatization and the contraction of social protection, create new inequalities. Types of Societies Before moving on to consider the evidence, we need to clarify the core typologies used to classify types of societies and states. Overall, for the global comparison 191 nation-states were classified according to levels of societal modernization. The Human Development Index produced annually by the United Nations Development Program (UNDP), provides the standard 100-point scale of societal modernization, combining levels of knowledge (adult literacy and education), health (life expectancy at birth), and standard of living (real per capita GDP). This measure is widely used in the development literature, and it has the advantage of providing a broader and more reliable indicator of societal well-being than monetary estimates based on levels of affluence.37 Using the 1998 Human Development Index (HDI), postindustrial societies are defined as the 20 most affluent states around the world, with an HDI score over .900 and mean per capita GDP of $29,585. Industrial societies are classified as the 58 nations with a moderate HDI (ranging from .740 to .899) and a moderate per capita GDP of $6,314. Lastly, agrarian societies are the 97 nations with lower levels of development (HDI of .739 or below) and mean per capita GDP of $1,098.38 22 I. The Causes of the Rising Tide table 1.2. Indicators of societal modernization % GNP Life % Gross from % Urban Expectancy % Adult Educational Type of Services, Population, (Years), Literacy, Enrollment Society 1998 2000 1998 1998 Ratio, 1998 Postindustrial 65.5 79-1 78 99.0 93-5 Industrial 59.2 67.2 73 93-3 75-3 Agrarian 45.6 40.9 59 67.0 54.1 All 51-7 53.8 66 79.2 65.5 Note: Mean social indicators in 172 nations. See Appendix A for the classification of nations and Appendix B for concepts and measures. Source: Calculated from UNDP. UNDP Human Development Report 2000. New York: UNDP/Oxford University Press. To see how far this classification predicts broader patterns of development, Table 1.2 illustrates some of the contrasts in the most common indicators of social well-being. The classic definition of postindustrial societies emphasizes the shift in production from fields and factories to the service sector. Almost two-thirds of GNP in the societies classified as postindustrial derives from the service sector, but this figure falls to only 45 percent in agrarian societies. Table 1.2 shows how levels of urbanization, literacy, education, and life expectancy systematically vary across classifications of different types of society. Perhaps the clearest contrast is in life expectancy: a person living in the average postindustrial society can expect to live seventy-eight years, as compared to only fifty-nine years in agrarian societies. Table 1.3 summarizes two basic indicators of change in levels of societal development from 1980 to 1998. The Human Development Index shows that there have been some gains in all societies during the last twenty years. Nevertheless, the disparities in levels of human development between societies currently classified as postindustrial and agrarian have scarcely closed at all. The per capita GDP in postindustrial societies has grown far faster than that of industrial societies, and agrarian nations have made no gains. In the next chapter, this initial classification is discussed further, and comparisons are drawn between this measure of societal modernization and the UNDP Gender-related Development Index (GDI), which takes into account the overall level of societal modernization and the disparities that can exist between women and men even in Introduction table 1.3. Trends in societal modernization, 19b, yu Human Development Index Per Capita GDP Type of Society 1980 1998 Change 1980 1998 Change Postindustrial Industrial Agrarian All .860 .742 .486 .625 .918 .802 .564 .681 +.058 +.060 +.078 +.056 20,932 6,155 1,099 5,506 29,585 6,314 1,098 6,162 +8,653 +159 — 1 2,565 Note: Real per capita GDP is measured in U.S. dollars estimated at purchasing power parity. See Appendix A for the classification of nations and Appendix B for concepts and measures. Source: Calculated from UNDP. UNDP Human Development Report 2000. New York: UNDP/Oxford University Press. affluent societies, as well as the Dikkstra and Hanmer Relative Status of Women (RSW) index, which measures gender equality within Countries.39 Type of States Over the years, numerous attempts have been made to develop effective measures of a given society's level of democracy. It should be noted that alternative measures emphasize different components, and that all the alternative indices suffer from certain conceptual or methodological flaws. Nevertheless, one recent review concluded that, despite these differences, in practice there is considerable similarity in the rank order correlations of nations across different indices.40 The Gastil index, a seven-point scale used by Freedom House, has become widely accepted as a standard measure providing a basic classification of political rights and civil liberties. We adopt this measure because it has the advantage of comprehensive coverage, including all nation-states and independent territories around the globe, as well as the ability to be used for time-series analysis, since the index has been published every year since the early 1970s. We have reversed the Gastil scale in the analysis for ease of interpretation, so that a higher score represents higher levels of democracy. We are also interested in how long democracy has endured in given societies. To obtain a measure of length of democratic stability, we use the annual Freedom House ratings produced from 1972 to 2000.4I We define as older democracies the thirty-nine states 2-4 I. The Causes of the Rising Tide table 1.4. Classification of societies and states in the pooled World Values Surveys /European Values Surveys, 1981-2001 Type of State Type of Society Older Democracy Newer Democracy Semi-democracy Non-democratic Total Postindustrial Industrial Agrarian Total 21 3 1 25 16 3 19 10 9 19 3 8 u 21 32 21 74 Note: The number of nations in each category. For details about the classifications, see Appendix A. around the world with at least twenty years' continuous experience of democracy from 1980 to 2000 and a Freedom House rating of 5.5 to 7.0 in the most recent estimate. We classify as newer democracies the forty-three states with less than twenty years' experience with democracy and a current Freedom House rating of 5.5 to 7.0. Another forty-seven states were classified as semi-democracies (Freedom House describes them as "partly-free"; others call them "transitional" or "consolidating" democracies); these states have been democratic for less than twenty years and have current Freedom House ratings of 3.5 to 5.5. Non-democracies are the remaining sixty-two states, with a Freedom House score for 1999-2000 from 1.0 to 3.0; they include military-backed dictatorships, authoritarian states, elitist oligarchies, and absolute monarchies. Appendix A lists the classifications of nations used throughout the book, based on these measures. Clearly there is considerable overlap between human and democratic development at the top of the scale; many older democracies are also affluent postindustrial societies. But the pattern of states among industrial and agrarian societies shows a far more complex pattern, with newer democracies, semi-democracies, and non-democracies at different levels of socioeconomic development. The World Values Survey The analysis of cultural attitudes is based upon the World Values Surveys (WVS), a global investigation of socio-cultural and political change. The study has carried out representative national surveys of Introduction 2-5 the basic values and beliefs of publics in n„*„ man seventy nation-states on all six inhabited continents (see Figure 1.1), containing in total 4.7 billion people or over 80 percent of the world's population. It builds on the European Values Surveys, first carried out in twenty-two countries in 1981. A second wave of surveys, in forty-three nations, was completed in 1990-91; a third wave was carried out in fifty-five nations in 1995-96; and a fourth wave with fifty-five nations took place in 1999-2001 (see Table A.2).42 The pooled survey used in this book includes almost a quarter-million respondents, facilitating subgroup analysis even for minority groups. The survey includes some of the most affluent market economies in the world, such as the United States, Japan, and Switzerland, with per capita annual incomes as high as $40,000 or more, together with middle-level industrializing countries, such as Taiwan, Brazil, and Turkey, and poorer agrarian societies, such as Uganda, Nigeria, and Vietnam, with per capita annual incomes of $300 or less. Some smaller nations, such as Malta, Luxembourg, and Iceland, have populations below one million; at the other extreme, almost one billion people live in India and well over one billion in China. The pooled survey with all waves contains older democracies such as Australia, India, and the Netherlands; newer democracies such as El Salvador, Estonia, and Taiwan; semi-democracies such as Russia, Brazil, and Turkey; as well as eleven non-democracies, exemplified by Zimbabwe, Pakistan, and Egypt. The transition process also varies markedly. Some nations have experienced a rapid consolidation of democracy during the 1990s; today the Czech Republic, Latvia, and Argentina currently rank as high in political rights and civil liberties as nations with long traditions of democracy such as Belgium, the United States, and the Netherlands.43 The survey includes some of the first systematic data on public opinion in many Islamic states, including Jordan, Iran, Algeria, Egypt, Indonesia, and Morocco. The most comprehensive coverage is available for Western Europe, North America, and Scandinavia, where public opinion surveys have the longest tradition, but countries are included from all world regions, including five sub-Saharan African nations and six Middle Eastern states. The four waves of this survey took place from 1981 to 2001, although the same countries were not always included in each wave, so comparisons over the full period can be carried out only in twenty societies. Data drawn from the Eurobarometer surveys, conducted biannually Introduction 27 since 1970, and from the Political Action _„„, ui tne mid-1970s facilitates longer-term comparisons. The Plan of the Book Based on these considerations, Chapter 2 goes on to analyze indirect attitudinal evidence of support for gender equality, including comparisons among different types of societies. With remarkable consistency, we find that the publics of richer, postindustrial societies are much more likely to support gender equality than the publics of agrarian or industrial countries. Intergenerational differences, which are largest in postindustrial societies and relatively minor in agrarian societies, suggest that the former are undergoing intergenerational changes. Women in postindustrial societies, in particular, are deeply divided by generation in their support for gender equality. Lastly, support for gender equality in the political sphere, in the workplace, and in the home is also explained by many of the standard factors commonly associated with cultural shifts, including education, religiosity, marital status, and postmaterialism. Chapter 3 considers the role of religion in more depth. In particular, the process of societal modernization is path-specific and is conditioned by the cultural heritage and structural context of a given society. This chapter demonstrates that the cross-sectional differences in support for gender equality vary even among societies at similar levels of human development, being shaped by factors such as the strength of religiosity and the type of religious values. Multivariate analysis probes these factors in more depth, and the chapter considers whether an Islamic religious heritage is the most powerful barrier to change. Part II of the book examines the political consequences of the rising tide of gender equality. Chapter 4 examines evidence of the shift from the traditional to the modern gender gap in voting behavior. The chapter compares cross-national support for parties across the left-right spectrum among women and men and considers how far these differences reflect the same modernization and cultural factors that shape attitudes toward traditional gender roles. The study also examines generational patterns in the size of the gender gap. One of the most intractable problems of gender equality concerns continuing male predominance in traditional political elites, such 28 L The Causes of the Rising Tide as parties and parliaments. Chapter 5 goes on to compare gender differences in three dimensions of political participation: traditional activism via elections and parties; civic activism through voluntary organizations, new social movements, and community associations; and protest activity, such as signing petitions and taking part in boycotts and demonstrations. The study shows how the major forces of modernization and cultural heritage affect these differences. Chapter 6. Do similar factors lead to the familiar gender gap in political leadership, in terms of women as heads of state, cabinet ministers, and parliamentary representatives? This chapter compares women and men in national parliaments around the world. The analysis considers the reasons for the persistence of gender differences and the roles of societal modernization and cultural legacies in explaining levels of female representation. Chapter 7. The conclusion examines how far attitudes toward gender equality form part of a larger cultural shift towards self-expression values. The conclusion draws together the major findings from all of the chapters and considers their implications for the transformation of women's and men's lives and for cultural change worldwide. 2 From Traditional Roles toward Gender Equality Developmental theory is based on the assumption that traditional societies are characterized by sharply differentiated gender roles that discourage women from working outside the home. An extensive literature in demography, sociology, anthropology, and social psychology has documented the familiar yet profound transformation of sex roles associated with the process of societal modernization.1 Virtually all preindustrial societies emphasize childbearing and child rearing as the central goal for women and their most important function in life, along with tasks like food production and preparation at home; jobs in the paid workforce are predominately male. In postindustrial societies, gender roles have increasingly converged because of a structural revolution in the paid labor force, in educational opportunities for women, and in the characteristics of modern families/ In most affluent countries, people are marrying later than in previous generations and having fewer children.3 A rapid increase in premarital cohabitation is challenging the once-privileged position held by marriage. More and more women, especially those who are married, have entered the paid labor force, creating the transition from male breadwinner to dual-earning families.4 Although the gender gap in rates of economic participation is narrowing, women's and men's roles in the labor force continue to differ. Women still have to juggle the demands of family responsibilities and market work, and they hold different jobs than men do, often with lower status and rewards. These social trends raise questions about long-established moral values and attitudes toward the 29 3° I. The Causes of the Rising Tide family and gender roles that were once taken for granted. Traditional family values have by no means disappeared, but they appear to be under greater strain in postmodern societies. Not all consequences of these sweeping developments can be examined here, but this account leads to four major predications that are open to empirical investigation - namely, we expect to find systematic differences in cultural indicators of gender equality: • between societies based on their level of economic development, • within societies based on generational cohorts, • between women and men, and • within societies based on structural and cultural factors such as education and class. In order to examine these propositions, this chapter analyses indirect attitudinal evidence, including comparisons among different types of societies. It demonstrates that richer, postindustrial societies support the idea of gender equality more than agrarian and industrial societies. Intergenerational differences, which are largest in postindustrial societies and relatively minor in agrarian societies, suggest that the former are undergoing intergenerational changes. Lastly, support for gender equality in the political sphere, in the workplace, and in educational opportunities is strongly related to patterns of education, religiosity, marital status, and postmaterialism, following familiar patterns of cultural attitudes. Measuring Attitudinal Support for Gender Equality How do we best measure attitudes toward gender equality? If this is a multidimensional phenomenon, then inconsistent trends could be apparent in different arenas, such as the home, the workforce, and the public sphere. The entry of married women into the paid labor market, for example, which expands pooled household incomes, may prove more socially acceptable to men than equality in the division of common domestic chores, such as routine house care and care of children and the elderly.5 Women working in sectors such as education, health, and voluntary organizations, reflecting traditional sex roles of women as caregivers, may encounter fewer barriers than those challenging conventional sexual stereotypes in military, political, and From Traditional Roles toward Gender Equal 31 religious institutions. Equal opportunity pi -„.^ung common classical liberal beliefs may prove more popular than strategies designed to achieve affirmative action, gender parity, or positive discrimination for women.6 Feminist philosophy contains multiple complex strands of thought, as reflected in deep-seated divisions among socialist, liberal, and cultural perspectives within the women's movement. Before examining the evidence, we first need to establish if attitudes toward gender equality fall into a single coherent ideological dimension. Since the early 1970s, an extensive literature in social psychology has developed and tested reliable and valid multidimensional scales of attitudes toward the division of sex roles in the home and workplace, including attitudes toward gender equality, feminism, the status of women, and support for the women's movement.7 Although useful, these comprehensive scales have usually been tested among small convenience samples of college students, often in the United States, restricting the generalizability of the results outside of that particular context. Cross-national comparisons have commonly used a more limited range of survey items available from sources such as the Eurobarometer and the International Social Survey Programme.8 This study develops a Gender Equality Scale by combining a battery of five items from the pooled 1995-2001 World Values Surveys / European Values Surveys. • MENPOL Q118: "On the whole, men make better political leaders than women do." (Agree coded low) (1990-2001 WVS/EVS). • MENJOBS Q78: "When jobs are scarce, men should have more right to a job than women." (Agree coded low) (1990-2001 WVS/EVS). • BOYEDUC Q.119: "A university education is more important for a boy than a girl." (Agree coded low) (1990-2001 WVS/EVS). • NEEDKID Quo "Do you think that a woman has to have children in order to be fulfilled or is this not necessary?" (Agree coded low) (1981-2001 WVS/EVS) • SGLMUM Q112 "If a woman wants to have a child as a single parent but she doesn't want to have a stable relationship with a man, do you approve or disapprove?" (1981-2001 WVS/EVS) These five items are similar to those commonly contained in the more comprehensive psychological scales of gender equality, tapping 32 /. The Causes of the Rising Tide table 2.1. Factor analysis of the Gender Equality Scale (five-item) Component Men make better political leaders than women. .710 Men should have more right to a job than women. .672 University education is more important for a boy. .641 Necessary for woman to have children to be fulfilled .556 Woman wants to have children as single parent .395 % of total variance 36.6 Note: Principal component factor analysis. Source: Pooled WVS/EVS, 1995-2001. attitudes toward politics, the workforce, education, and the family. Three items use statements with Lickert-style four-point agree-disagree responses, while two use dichotomies; these items were all recoded so that higher values consistently represent greater support for gender equality. Principal component factor analysis revealed that all five items tap a single dimension (see Table 2.1), with a Cronbach's Alpha of 0.54.9 The Gender Equality Scale was summed across the items and standardized to 100 points for ease of interpretation. The full five-item scale is available for the 1995-2001 waves of the WVS/EVS in sixty-one societies. Two items (NEEDKID and SINGLEMUM) are available in twenty societies for time-series analysis over all four waves since the early 1980s. Cross-national Comparisons If there are coherent and predictable value shifts associated with societal modernization, then the most egalitarian attitudes toward the division of sex roles should be found in the most affluent societies. This is certainly what we would expect given other major cultural shifts documented in previous work, including the transition from traditional to secular-rational values and the transition from survival to self-Expression values.10 The distribution of sixty-one nations on the gender equality scale for 1995-2001, shown in Figure 2.1, provides preliminary support for the proposition that attitudes toward traditional or egalitarian roles for women and men vary systematically according to levels of economic development. The countries ranking as most egalitarian include some of the most affluent in the world, such as Finland, Venezuela Slovenia Chile B Uruguay ■ Croatia I Mexice I Macedonia I El Salvador I S.Africa I Lithuania I Serbia I Brazil I Bulgaria I Czech I Hungary ■ Albania ■ Latvia I Bosnia I Estonia I Montenegro I Russia I Vietnam ■ Romania I Zimbabwe I Japan I Uganda I Slovakia ■ Ukraine ■ Poland I Tanzania B China I Belarus I India I Turkey Moldova Taiwan I Georg a I Azerbaijan I Philippine I Iran I Morocco I Egypt I Bangladesh I Jordan I Armenia I Nigeria — W.Germany ■H Canada _K Norway ■m Spain I E.Germany I New Zealand I U.S. I Colombia BB Australia I Argentina Peru Dom.Repub 45 60 75 90 figure 2.1. Gender Equality Scale by nation. Combined 100-point scale of the following 5 items: MENPOL Q118: "On the whole, men make better political leaders than women do." (Agree coded low); MENJOBS Q78: "When jobs are scarce, men should have more right to a job than women." (Agree coded low); BOYEDUC Q119: "A university education is more important for a boy than a girl." (Agree coded low); NEEDKID Quo "Do you think that a woman has to have children in order to be fulfilled or is this not necessary}" (Agree coded low); SGLMUM Q112 "If a woman wants to have a child as a single parent but she doesn't want to have a stable relationship with a man, do you approve or disapprove?" (Disapprove coded low). Source: WVS/EVS, pooled 1995-2001. 33 34 I. The Causes of the Rising Tide Sweden, Canada, and West Germany, each scoring over 80 percent. Many moderate-income industrialized nations, such as Brazil, Mexico, and Bulgaria, rank in the middle of this scale, while some of the poorest societies fall at the bottom, including Nigeria, Armenia, Morocco, Bangladesh, and Jordan, all ranked with GDP per capita of $1,500 or less in 1998 and gender equality scores below 60 percent.11 Overall, the mean score on the 100-point scale was 80 percent for postindustrial societies, compared to 68 percent for industrial nations and 60 percent for agrarian societies. To analyze the pattern more systematically, the degree of support for gender equality in each country was compared to levels of economic development, measured by logged per capita GDP (in purchasing power parity U.S. dollars). The result, shown in the scattergram in Figure 2.2, confirms the consistency and strength of the association between gender equality and development (R1 = 0.54). All of the postindustrial 90 80 I 70 "T LO ® CO O CO >. To O" LU 60- 50- CD ■o C CD O 40 Swe Col „ Can • Nor %* »m "AustL^ DomRPenv m m ,t Chil Arg JUS MaEISaLi,h sAÄ Alb H B"l f". ^zech >,. Bos Lat s' "-i, ékjir Est ä; Tanz W 2"l)kr ti * India CtaI» s Mol»/' ™3^ & Pol ,___ Bela Turí* Jap 42S Geo Nigeria^» i> Arn?nfan Bng s T-----------1------1-----1—i----------- EgMMr T-----------------1----------1------1-----1-------- ---------1-----------------1----------1—1— %%•%% °o °o°o°o Type of Society a Agrarian '•-'■ Industrial * Postindustrial Rsq = 0.5401 %ýoMpn. %U°$%o0 Logged per capita GDP 1998 ($US PPP) figure 2.2. Gender Equality Scale and economic development. G ender Equality Scale: see Figure 2.1. Logged per capita GDP 1998: see UNDP. Human Development Report 2.000. New York: Oxford University Press/UNDP. Source: Pooled WVS/EVS, 1995-2001. From Traditional Roles toward Gender Equa, 35 nations, with the important exception of J .istered in the top right-hand corner, as most strongly in favor of gender equality and also most affluent. A range of industrial societies cluster in the middle of the distribution, with Latin American nations such as Colombia and Venezuela more egalitarian in their attitudes toward sex roles than post-Communist societies such as Poland, Belarus, and Russia. This suggests that other factors are also influential, beyond economic development, such as the historical role of the state under Communist regimes and the religious legacy that predominates in certain areas of the world. Lastly, agrarian societies proved to be least egalitarian, although again there were important differences among these poorer nations that deserve further exploration in the next chapter, such as the contrasts apparent between Vietnam and Bangladesh, or between India and Azerbaijan. These results suggest that although economic prosperity is one of the factors most strongly associated with the existence of an ethos of equality between women and men in a society, it is far from the whole story. Amartya Sen has drawn an important and influential distinction between the acquisition of income and wealth and the broader notion of human development, with the latter containing many other important indicators of societal well-being, such as the ability to enjoy clean water and sanitary facilities, to obtain sufficient nutrition, and to get adequate shelter and warmth, as well as conditions of freedom from conflict and crime and social exclusion, all of which shape prospects for survival into old age.12 Economic growth can influence these conditions, but even in affluent nations there can still be many pockets of social inequality and inadequate safety nets. There are sharp contrasts between high-growth economies such as South Korea that have had considerable success in raising the quality of life, and others such as Brazil that have a history of severe social inequality, unemployment, and neglect of public health care. To explore these contrasts further, the UNDP defines and measures human development by factors such as longevity, education, and literacy as well as GDP; they have collected a wide variety of statistics that can be used to assess and compare progress in societal modernization during the 1990s.13 Many of these indicators, such as estimates of the average daily supply of calories, per capita electricity consumption, and the percentage of GNP derived from agriculture, are essentially gender-neutral. Others are more 36 I. The Causes of the Rising Tide closely related to the conditions of women's lives and inequalities between the sexes, such as the ratio of women to men at different educational levels, the proportion of births to young mothers, and the contraception rate. The comparison of many nations in Table 2.2 shows that across a wide range of selected indicators, the attitudes on the gender equality scale are strongly and significantly related to both the societal and the gender-related indicators, in the predicted direction. Of course, there could be an interaction effect at work with some of these indicators, if cultures favorable to gender equality are the ones that most facilitate equal opportunities for women in education or employment. But even those developmental indicators demarcating poorer nations that are not directly related to gender relations, such as basic levels of food aid, energy use, and the debt service ratio, prove to be strongly associated with egalitarian beliefs regarding sex roles. This supports the contention that societal modernization leads to more secure lives, primarily through increasing personal incomes and wealth and facilitating effective public sector and nonprofit programs for social protection, which in turn gradually create a climate conducive to more liberal views of social roles, including those determined by sex, class, and ethnicity. When life is no longer "nasty, poor, brutish and short," restricted by widespread fears and insecurities based on life-threatening challenges, then women and men gradually develop greater willingness to adopt interchangeable roles within the family and workforce. Correlations cannot prove causation, and indeed, there could in theory be an alternative 'X' factor (as there is in Weberian theories of the Protestant ethic) that simultaneously drives both support for gender equality and levels of societal development.14 But it is clear from this wide range of evidence from many nations, and consistent with our theoretical interpretation, that where societal modernization and human development have progressed furthest, traditional conceptions of a strict demarcation between the roles of women and men have broken down most fully. Generational Change and Cohort Analysis So far we have established differences between rich and poor nations, but not whether attitudes toward women have become more liberal From Traditional Roles toward Gender Equal 37 table 2.2. Gender Equality Scale and indicators of sc . .ization Year Correlation Sig. Number of Nations Societal indicators Per capita GDP (PPP $US) Human Development Index (HDI) % Urban population % GNP from services % GNP from agriculture Average life expectancy (years) Daily supply of calories Per capita commercial energy use (oil equivalent) Per capita electricity consumption (kw-hours) Debt service ratio Dependency ratio (%) Food aid in cereals (thou, metric tons) Drug crimes (per 100,000 people) Level of democratization (Freedom House Index) Gender-related indicators Gender-related development index (GDI) Gender empowerment measure (GEM) Female administrators and managers (% of total) Female professional and technical workers (% of total) Female primary net educational enrolement, as % of male enrollment Female secondary net educational enrolement, as % of male enrollment Female tertiary students, per 100,000 women Female adult literacy rate, as % of male rate Births to mothers under 20 (%) Contraceptive prevalence rate (%) 1998 .608 .000 55 1998 .666 .000 55 1998 .538 .000 51 1998 .389 .000 47 1997 -.587 .000 45 1998 .505 .000 52- 1996 .482 .000 51 1996 .611 .000 48 1996 1997 .635 .709 .000 51 1997 .415 .007 41 1997 —.271 .054 51 1994-5 —.609 .003 22 1994 .616 .000 32 2000 •537 .000 58 .000 52 1997 .843 .000 41 1997 •493 .001 41 1997 .420 .006 41 1997 .556 .000 42 1997 .566 .000 41 1996 .623 .000 34 1997 .677 .000 37 I99I-7 -.469 .003 39 1990-8 .498 .001 39 Noře; The figures represent correlation coefficients between national scores on the Gender Equality Scale (from the pooled WVS/EVS, 1995-2001) and selected indicators of societal modernization, without any controls. For the 100-point standardized Gender Equality Scale (5-item), see Table 2.1. For the definitions, measures, and sources of the social indicators, see Appendix B. Source: UNDP. 2000. Human Development Report 2.000. New York: Oxford University Press/UNDP. 38 I. The Causes of the Rising Tide over the years. For indirect evidence of long-term trends, cohort analysis can be used to examine whether, as predicted, the younger generations are more in favor of equal roles in the home, the workplace, and the public sphere than their parents' and grandparents' generations. The theory of value change argued here suggests that secular social trends have only a glacial effect on cultural norms but that, through the socialization process, the conditions experienced during the formative years of childhood and early adolescence make an indelible impression on people. As a result, the values held in later life continue to be shaped by these seminal early experiences. Certain decisive historical events and common experiences also leave an imprint on a generation. Those growing up during the interwar era in Western nations experienced the dramatic collapse of stocks and savings, mass unemployment, and soup kitchens during the 1930s triggered by the Great Depression, followed by a military conflict that engulfed the world at the end of the decade. Given these conditions, the interwar generation in postin-dustrial societies is likely to prioritize materialist social goals, such as the importance of secure and full employment, low inflation, and the underlying conditions for economic growth, and to hold traditional views toward the division of household and parental responsibilities and support for authorities. By contrast, the postwar generation in these nations, coming of age during periods of unprecedented affluence, domestic peace, and social stability, are more likely to adhere to postmaterialist values, including equality between the sexes in sharing household tasks and equal opportunities in the labor market. Using only cross-sectional survey evidence, it is difficult to disentangle generational effects from life-cycle effects that may alter attitudes and values as people move from youth to middle age and then to retirement. The experience of education, entry into the labor force, child rearing, and old age can all be expected to shape beliefs about appropriate sex roles in the home and workplace. Cultural messages conveyed by the mass media and contact with organizations such as the women's movement can also color perceptions about appropriate attitudes and behavior for women and men. Significant changes in the lives of women and men may also generate a period effect, exemplified during the 1960s by the availability of safer contraception and wider access to abortion in many countries. Nevertheless, we assume that the acquisition of sex roles and core values of gender equality learned early in life in From Traditional Roles toward Gender Equat 39 90' ________ 1- Postindustrial ^-"""" O S 8° ^''' LO O) 05 jS* Industrial .,,-- y^ __........------'" iň 0) ffl 60 ■'" 0 CO ^-----——■ Agrarian & CO D O" en. Ol O ■0 c OJ CD 40 , \ ■*> ■*> '<>>> figure 2.3. Gender Equality Scale by cohort and type of society. Gender Equality Scale: see Figure 2.1. Source: Pooled WVS/EVS, 1995- the family, school, and community are part of the primary socialization process, so that the enduring values of different birth cohorts can be attributed mainly to their formative experiences in childhood and adolescence. In order to examine the evidence supporting these propositions, support for gender equality can be analyzed both by type of society and by ten-year birth cohorts. The results, presented in Figure 2.3, illustrate the substantial generational gaps in cultural values in affluent nations. In postindustrial societies, members of the interwar generations, born between the 1920s and the 1940s, prove far more traditional in their beliefs about the appropriate division of sex roles. By contrast, the postwar baby boom generations display far more egalitarian attitudes. This generation that grew up during the affluent fifties and came of age during the sixties and seventies witnessed the rise of the second-wave women's movement, expanded educational opportunities for women, and growing female participation in the paid workforce, in routine white-collar jobs, and in careers as managers, administrators, and professionals. Moreover, in these societies, after a rapid rise among the interwar generations, there is a plateau evident among the postwar generations, so that cohorts born in the 1950s, 1960s, and 1970s share similar egalitarian attitudes. The cultural changes in affluent nations 40 I. The Causes of the Rising Tide have therefore proved enduring and stable. Within postindustrial societies, there is a ten-point gap on the gender equality scale between the interwar and postwar generations. By contrast, the line illustrating cohort change in industrialized nations is far flatter, and there is a six-point gap between the pre-war and postwar generations. Lastly, in the poorer, agrarian societies there is an almost flat line across cohorts, except for the oldest group, and there is only a one-point gap between the interwar and postwar generations. But has this shift mainly affected women, or have men changed along with women? Some survey evidence comparing attitudes toward marriage and the family in Britain, Ireland, the United States, and West Germany found greater contrasts between men and women than across countries.15 Yet other studies have documented important generational shifts in male sex-role attitudes in America.16 To explore these issues, Figure 2.4 shows the generational gap for men and women in support for gender equality. This analysis confirms that the growing belief in equal sex roles evident in affluent nations is strongest among the postwar generation of women; nevertheless, the rising tide is not simply confined to this group, as in these nations younger men have also gradually come to favor equal opportunities for women and men. Interestingly, both women and men in affluent societies seem to have reached a "plateau" in support for gender equality, as there appears to be little difference between postwar baby boomers born in the late 1950s and subsequent cohorts. Most importantly, the patterns show that the key contrast, and indeed the growing gap in values and attitudes, is between types of societies rather than between women and men. This pattern is particularly marked on three issues - women's leadership in politics, equal opportunities in the paid workplace, and the need for women to have children in order to be fulfilled. These findings have implications that can be interpreted in two alternative ways. It is often assumed in the West that claims for women's rights are universal, and that men in patriarchal societies have restricted women's opportunities for secondary education, employment, and empowerment, and their rights to divorce, reproductive freedom, and property. In this perspective, the legal, political, and social system in traditional societies reflects and buttresses men's interests. This argument assumes that women in these countries, lacking empowerment and silenced in the public sphere, are unable to express their real Postindustrial Women /''" síäSÍSS í 5 g ä S s^s%% ^ Men -^ -fo_ -To. ■*> ■» ■¥? 'ß <& ď6 ■ '<& '% '% '% ' '\ ■W *ÍP Xí % °<9 '» 6 °fc figure 2.4. Societal differences on the Gender Equality Scale. 41 42 I. The Causes of the Rising Tide dissatisfaction with the status quo. In this view also, where women's voices are silenced, a coalition of international agencies such as the United Nations, working with multiple transnational feminist networks, women's organizations, and NGOs, can legitimately intervene on behalf of these women to demand that governments introduce equal opportunities. But the results of this analysis suggest an alternative proposition, namely, that common, deeply engrained belief systems in poorer developing societies mean that both men and women willingly adhere to the traditional division of sex roles in the home, family, and workplace. In this regard, men in these societies are not actively restricting and silencing women's demands; instead, both sexes believe that women and men should have distinct roles, as primary breadwinners and caregivers. It is true that due to the hegemonic grip of patriarchal ideology pervasive throughout such societies, and reinforced by the socialization process in the family, schools, religious organizations, and the media, women in these nations may be suffering from problems of "false consciousness" about their "real" interests, as many Western feminists might argue. Younger generations of educated professional women in these societies may become increasingly dissatisfied with traditional beliefs, creating coalitions with feminists elsewhere. Nevertheless, the evidence suggests that, if given a free choice, most women in these societies would still willingly opt for traditional roles in the home and family, leaving men predominant in the public sphere. In this regard, women may be prisoners, but they are willing prisoners, with the limits to their roles set by their socialization as much as by society's laws. Egalitarian ideas common among feminists in North America and Western Europe may eventually prove universally applicable, but these beliefs may be less widely shared than is commonly recognized. The evidence, therefore, suggests that in affluent Western nations more egalitarian beliefs have gradually become predominant among both women and men, reflecting the changed lifestyles and experiences of the postwar generation. In agrarian societies, by contrast, traditional attitudes continue to prevail among both men and women. This statement needs qualifying, since women favor gender equality slightly more strongly than men within every society. Nevertheless, it appears that, despite claims for the universalism of women's rights, there is a growing gap between the egalitarian beliefs and feminist values of Western societies and the traditional beliefs in poorer societies. Possible From Traditional Roles toward Gender Equai 43 consequences of this widening cultural clea explored fur- ther in the conclusion. Comparisons within Societies What individual-level factors best explain support for gender equality? Previous studies comparing feminist attitudes in the fifteen EU member states have reported that at the individual level, the social background of women has often proved important, including their labor force participation, age, education, religiosity, and partisanship.17 A regression model was used to examine predictors of support for the gender equality scale in the pooled WVS/EVS surveys across all fifty nations for 1995-2001, including social background variables that are often found to be associated with modern and traditional attitudes (age, gender, income, religiosity, work status, marital status, and children), as well as family savings, taken as a proxy measure of household economic security. Controls were included for the level of human development and democratization of society, and global regions were included as an indicator of residual cultural factors around the world. The results in Table 2.3 show that all of these factors proved to be significant, with coefficients in the expected direction. Among these variables, levels of human development proved to be one of the strongest predictors of attitudes. Comparing just the social and demographic factors, with societal controls, support for gender equality was strongest, as expected, among the younger generation, women, the better educated, and the less religious. But marked regional coefficients remained, suggesting that levels of societal modernization and the social backgrounds of respondents were insufficient to account for the remaining contrasts among those living in Asia and Central and Eastern Europe (who were less supportive of gender equality than average) and among Scandinavians (who were the most egalitarian). This suggests that social groups with similar characteristics living in different regions, such as professional women or older home workers in New York and New Delhi, or in Stockholm and Santiago, would still display divergent values. The next chapter will go on to explore these contrasts in more detail, including the imprint of cultural legacies produced by different religious traditions, state regimes, and national histories. 44 I. The Causes of the Rising Tide table 2.3. Social characteristics and support for the Gender Equality Scale Standardized Unstandardized Standard Beta Coefficients (B) Error Coefficients Sig. (Constant) 44.589 .875 .000 Societal development Human Development 37.176 •947 .234 .000 Index, 1998 Level of democratization, 1.416 .073 .104 .000 1999-2000 Social background Age (years) -.147 .004 -.139 .000 Gender (men = 1, — 5.102 .138 -•137 .000 women = 0) Educational 2-947 .096 .116 .000 (three categories) Income .000 .000 .021 .000 Frequency of church -.817 .036 —.090 .000 attendance (seven-point scale) Work status (paid .551 .094 .022 .000 work = 1, else = 0) Married or living as -.800 .154 —.021 .000 married (1, else = 0) With at least one child — 1.125 .235 —.020 .000 (1, else = 0)) Family savings in -.424 .073 —.022 .000 past year (four-point scale) Region Asia — 11.506 .224 —.246 .000 Central and — 10.121 .213 -.234 .000 Eastern Europe Middle East — 8.991 .298 —.130 .000 North America -1.979 .278 —.032 .000 Africa -.830 .381 —.011 .029 Western Europe 1.474 •394 .015 .000 Scandinavia 1.988 •379 .023 .000 Adjusted R2 .28 Note: OLS Regression models with the Gender Equality Scale, 1995-zooi (five the dependent variable. Regional dummy variables (0/1) exclude South America coding of variables, see Appendix B. Source: Pooled WVS/EVS, 1995-2001. item) as . For the From Traditional Roles toward Gender Equa, 45 table 2.4. Mean scores on the Gender Equality Agrarian Industrial Postindustrial Total All 60 68 81 67 Gender Women 65 70 81 70 Men 59 66 77 64 Diff. 6 4 4 6 Education High 63 69 83 69 Moderate 60 69 80 68 Low 56 65 76 63 Diff. 7 4 7 6 Age group Under 30 63 72 84 69 30-59 61 67 81 67 60+ 59 62 71 64 Diff 4 10 13 5 Postmaterialism (four-item) Postmaterialist 69 75 85 80 Mixed 63 68 78 68 Materialist 59 63 72 62 Diff. 10 12 J3 18 R's occupational class (V230) Manager/professional 63 69 80 70 Lower middle 66 71 81 73 Skilled working 61 66 79 67 Unskilled working 61 65 78 65 Diff. 2 4 2 5 Work status Women in paid work 64 70 82 71 Men in paid work 59 66 77 65 Women looking 61 70 76 66 after home Diff. among women 3 0 6 6 Marital status Separated 69 74 83 75 Cohabiting 66 71 85 73 Divorced 69 70 82 73 Single 66 72- 84 71 Married 61 66 77 66 Widowed 62 63 71 64 Diff. 7 11 12 13 (continued) 46 I. The Causes of the Rising Tide table 2.4 (continued) Agrarian Industrial Postindustrial Total Children No children 65 72- 84 70 At least one child 62 67 79 67 Diff. 3 5 4 3 Note: For coding, see Appendix B. For the 100-point Gender Equality Scale, see Table 2.1. Source: Pooled WVS/EVS, in sixty-one societies, 1995-2001. Table 2.4 breaks down the mean gender equality scale by social group and type of society. In addition, the influence of postmaterialism was also examined, as one of the best-established indicators of cultural change, using the standard four-part scale that divides people into materialists, mixed, and postmaterialists.18 The results confirm our expectations, with the highest scores on gender equality (over 80 percent) found in postindustrial societies among women, those under thirty, those who were unmarried or widowed, those without children, and among adherents of postmaterialist values. Among affluent nations, the generation gap proved a stronger predictor of egalitarian attitudes than sex, class, or education, strongly supporting the thesis of long-term secular cohort change in these fundamental values. Conclusions Early accounts emphasized the importance of economic growth for promoting gender equality, and subsequent decades saw a strong focus on establishing effective legal, social, and political rights for women. Cultural accounts have often claimed that values are equally important, but it has been difficult to examine that proposition systematically, in large part because, without cross-national surveys, the predominant approach has been to compare social norms and behavior through qualitative techniques such as personal interviews, participant observation, and textual exegesis, usually covering one or a few societies. Cultural relativists argue that such an approach is appropriate, because subtle differences can be detected in attitudes toward women even among relatively similar nations with different cultural histories, such as Germany and France, or Canada and the United States. This From Traditional Roles toward Gender Equa 47 chapter has compared systematic survey ev iny nations at different levels of societal modernization. Survey questions may fail to catch some of the more nuanced aspects of complex social relations; nevertheless, the WVS data provides a broad map of the state of beliefs and values about sex roles in many societies around the globe. The results, based on analysis of the 100-point Gender Equality Scale, suggest five main conclusions: 1. Far from being a random distribution, attitudes toward gender equality form coherent and predictable patterns. 2. In particular, there are clearly established contrasts among countries at different levels of societal modernization, with agrarian nations the most traditional in their perceptions of sharply divided sex roles, industrial societies in the early stages of transition, and postindustrial societies the most egalitarian in their beliefs about the roles of women and men. 3. Moreover, this is not just a matter of economic development, because a wide range of non-gender-related indicators of human development, from levels of energy use to average life expectancy, are equally good predictors of support for gender equality. 4. The analysis of generational differences also showed predictable patterns, with younger generations in postindustrial societies being far more egalitarian than their parents and grandparents. This generational change was less evident in industrialized nations, and within poorer agrarian societies there was no evidence of significant generational shifts. 5. Within societies, there are significant differences between women and men, but in postindustrial nations younger men have also shifted their values along with younger women. Support for gender equality was also stronger among the well educated, the less religious, the unmarried, and among postmaterialists. But perhaps the most important finding is that the gap that has emerged between traditional agrarian societies and egalitarian postindustrial societies is far greater than the gap that exists between women and men within each type of society. Despite the role of human development, attitudes toward gender equality were still found to vary even among societies at similar levels of 48 I. The Causes of the Rising Tide human development, according to factors such as religious traditions, type of state, and level of democratization. Among rich nations, there are substantial contrasts between Japan and Norway, for example, just as there are between poorer societies such as Jordan and Columbia. We will attempt to understand these factors in more depth in the next chapter. 3 Religion, Secularization, and Gender Equality We have established the existence of systematic and predictable differences in cultural attitudes toward gender equality that vary across nations according to their level of human development, within societies based on generational differences, and among social groups according to factors such as education and marital status. Building on this foundation, this chapter demonstrates that the process of societal modernization in any given society is conditioned by cultural legacies and religious traditions. We hypothesize that (i) a process of secularization has gradually accompanied societal modernization, weakening the strength of religious values in postindustrial societies, particularly among the younger generation, and fuelling the rising tide of gender equality. (2) Postindustrial societies have experienced a parallel liberalization of attitudes toward sexuality, exemplified by the issues of abortion, homosexuality, prostitution, and divorce. (3) Yet religiosity continues to exert a strong influence on social norms about the appropriate division of sex roles in the home, the workforce, and the public sphere, especially in agrarian societies. (4) Moreover, attitudes toward women vary among adherents of different religious sects and denominations; in particular, an Islamic religious heritage is one of the most powerful barriers to the rising tide of gender equality. The literature suggests multiple reasons why religion can be expected to exert a major influence over prevalent attitudes and practices regarding sex roles. 49 SO /. The Causes of the Rising Tide • An extensive body of work in sociology, social psychology, anthropology, and theology studies suggests that religion has functioned as one of the most important agencies of socialization determining social norms and moral values with regard to gender equality in all societies, and influencing support for feminism and attitudes toward the second-wave women's movement. The role of the church in this process is similar to that of other important socialization agencies, including the family, community, school, workplace, and the mass media.1 • Religious organizations, particularly the Catholic Church and the evangelical movement among fundamentalist Christians in the West, and Islamic fundamentalist leaders in Muslim nations, have often actively sought to reinforce social norms of a separate and subordinate role for women as homemakers and mothers, buttressing traditional policies and the legal framework regulating marriage and divorce, abortion and contraception, family and childcare policy.2, • Research on electoral behavior and public opinion, influenced by Duverger's seminal work on women and politics in Western Europe in the 1950s, has long regarded women's greater religiosity as an important influence on patterns of partisanship and voter choice, including greater female support for parties of the center-right, such as Christian Democrats and Conservatives.3 • Studies of political representation, legislative elites, and leadership recruitment have established that the type of religious culture acts as an important contextual factor inhibiting women's entry into elected office. In particular, recent cross-national studies have found that fewer women enter legislatures in predominately Catholic and Islamic societies, controlling for many other common factors such as levels of economic development, democratization, and types of electoral systems.4 • Lastly, in recent years considerable research in the human rights literature has focused attention on the problems facing women in Muslim societies, and the question of whether rights such as citizenship can be regarded as universal if they conflict with pervasive religious beliefs, social norms, and Islamic law governing the role of women in the Middle East.5 Religion, Secularization, and Gender Equality 51 If traditional religious beliefs and practices are eroding in many affluent nations, with the growing secularization of the modern world, this could help explain the increasing support for gender equality in the home and workplace that we have already observed in postindustrial nations, especially among younger generations.6 This is a plausible proposition that deserves analysis; but so far, little systematic cross-national survey data has been available to examine how religious beliefs and practices vary around the globe, or how they may ebb and flow in response to the processes of societal modernization and broader shifts in the cultural Zeitgeist. Moreover, even the basic contention of growing secularization remains under challenge. Some divinity scholars point to evidence of religious revivals and countersecular movements, such as the vigorous resurgence of Orthodox Judaism in Israel, as well as to fundamentalist Islamic movements in many countries in the Middle East and North Africa and the development of new religious movements and evangelical revivals in the West.7 In particular, Samuel Huntington has claimed that following the end of the Cold War, civilizational cultures, based largely on religious values, are playing a larger role in domestic and international conflict.8 We therefore need to compare the evidence to see whether there has been growing secularization, as commonly assumed, before then examining the relationships among the strength of religious beliefs, the type of religious faith, and support for the values of gender equality. The Loss of Religious Faith? The theory that modernization has led to secularization was emphasized in the work of Max Weber and was popularized among sociological and theological writers during the 1950s and 1960s.9 At its simplest, secularization theories suggest that modernization leads to the decline of religious beliefs, as indicated by the erosion of church attendance, denominational allegiance, and faith in religious authorities; the loss of prestige and influence of religion's symbols, doctrines, and institutions; and the growing separation between church and state.10 Modernization theories suggest that growing levels of literacy and education, and wider sources of information, have strengthened rational belief in scientific knowledge, expert authorities, and technological know-how, 52 I. The Causes of the Rising Tide with priests, ministers, rabbis, and mullahs regarded as only one source of authority, and not necessarily the most important one, competing with the expertise of experts and professionals such as psychologists, physicists, and physicians.11 The evidence supporting secularization is strongest for Western Europe, where many people have ceased to be regular churchgoers outside of special occasions such as Christmas and Easter, weddings and funerals, a pattern particularly marked among the younger generations.11 Evidence from the Eurobarometer surveys conducted since 1970 suggests a consistent erosion in church attendance experienced across the fifteen Europe Union member states during the last three decades, with a fairly steep fall found in Belgium, Luxembourg, and the Netherlands and a relatively shallow decline over the years in Germany, France, and Ireland.13 Nevertheless, despite gradually emptying pews, the disparities in religiosity within European societies remain marked, depending upon historical traditions and the power of the church, producing contrasts such as the continuing hegemonic grip of the Catholic Church in Ireland and the far more secular society evident in Protestant Denmark. Even within Europe, however, a distinction needs to be drawn between behavioral indicators such as habitual attendance at church services, which has fallen, and religious values and beliefs, which may persist.14 Among affluent nations, the United States provides an interesting anomaly, with church attendance almost as high today among Americans as it was sixty years ago. According to Gallup polls, in 1939 about four out of ten Americans reported attending church or synagogue every week; almost the same proportion persisted, with minor fluctuations, in the most recent (2001) polls.15 Some erosion is evident in other indicators of religiosity in the Gallup series; for example, about two-thirds of Americans (65%) currently count themselves members of a church or synagogue, down from almost three-quarters (73%) in 1937. The salience of religion has also fallen slightly: today about two-thirds of Americans (64%) report that religion is "very important" in their own life, down from three-quarters (75%) in 1952.l6 Elsewhere in the world the picture remains complex, and it is difficult to establish reliable data on longitudinal trends and cross-national comparisons in churchgoing and religious affiliation. The World Christian Encyclopedia compares churches and religions around the globe and provides the most comprehensive estimates of secularization Religion, Secularization, and Gender Equality table 3.1. Factor analysis of the Strength of Religiosity Scale (six-item) Component Importance of God .883 Comfort in religion .846 Belief in God .836 Religious identity .797 Attend religious services .716 Life after death .653 % of total variance 62.8 Note: Principal component factor analysis. Importance: Vi 96. "How important is God in your life." (% "Very," scaled 6-10) Comfort: V. 19 7. "Do you find that you get comfort and strength from religion?" (%Yes) Believe: V191. "Do you believe in God?" (%Yes) Identify: V186. "Independently of whether you go to church or not, would you say you are... a religious person, not a religious person, or a convinced atheist?" (% Religious) Attend: V185. "Apart from weddings, funerals and christenings, about how often do you attend religious services these days?" {% Once a week or more) Life: V192. "Do you believe in life after death?" (% Yes) Source: WVS/EVS, pooled sample 1981-2001. during the twentieth century, based on an annual religious "mega-census" completed by ten million church leaders, clergy, and other Christian workers.17 The study suggests that the proportion of "non-religionists," defined as including agnostics, atheists, and other nonreligious groups, grew from an estimated 3.2 million in 1900 (0.2% of the globe's population) to about 19% in 1970 and peaked at 21% of the world's population in 1980, before falling back to 15% by 2000, following the collapse of Communism and the revival of organized religion in Central and Eastern Europe. Cross-national survey evidence is needed for more reliable comparisons, using multiple indicators of religiosity to take account of diverse religious practices among different sects and denominations. To develop a systematic Strength of Religiosity Scale, Table 3.1 presents the results of factor analysis using six indicators from the pooled WVS/EVS for 1981-2001, namely, the proportion of the population in different 54 I. The Causes of the Rising Tide mammmnmmiiMiMmBaBBfflaafflM^^^^^nw^^ai^^^nanM^aB^^n Jordan E3BBSBB33EBBBBBBB3BBBB3B33BBBBBBB3BB3B1 Egypt ■■■■^■^HHIHBHBBHHMBBHHHIB^^^MHBHHIHH^H^BH^B^H^iBII Bangladesh »^^^^^■■■■■^^■IMH^MMWMM——— Malta [ ^^^■H^^^a^^^^M^^^—rom^MM^^m Zimbabwe &B^^^B^«raBn—BaBBm^m«^^^^«i««ai^^^M^m^^Hn^«i^ Phiiipbines ■-■wiiiimi^^mBmiBBBmai^^n^n^^Bii^^an—^MBaaan^w^BnraiM Uganda 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 figure 3.1. Strength of Religiosity Scale. Note: For the items on the Strength of Religiosity Scale, see Table 3.1. Source: Pooled WVS/EVS, seventy-five societies, 1981-2001. societies: (i) who say that religion is "very important" in their lives, (2) who find comfort in religion, (3) who believe in God, (4) who identify themselves as religious, (5) who believe in life after death, and (6) who attend religious services regularly. All of these items tap values Religion, Secularization, and Gender Equality 55 and beliefs common throughout the world's religions, and they were carried in all four waves of the WVS, making possible comparison over time. Factor analysis showed that all of the items tapped a common dimension and formed a consistent and reliable Strength of Religiosity Scale (Cronbach's Alpha = 0.48). After recoding, the scale was standardized to 100 points, for ease of interpretation, where higher scores represent stronger religiosity. The comparison shows substantial contrasts in the strength of religiosity worldwide (see Figure 3.1). Low-income societies in Africa and Asia, exemplified by Nigeria, Uganda, the Philippines, and Zimbabwe, emerge as the most religious according to this scale, while at the other end of the spectrum, some of the post-Communist societies are among the most secular, including the Czech Republic and East Germany as well as postindustrial societies in Northern Europe, such as Sweden and Denmark. Nevertheless, although development is related to religiosity, there are many important exceptions to this pattern, above all the continuing strength of religiosity in the United States and, by contrast, the lower-than-expected levels in South Korea, Vietnam, and Tanzania. When broken down by type of society (see Table 3.2), agrarian table 3.2. Strength of religiosity by type of society (percent) Agrarian Industrial Postindustrial Total Believe in God Believe in life after death Religion "very important" Identify as religious Comfort from religion Attend religious service regularly 97 83 87 73 74 47 80 62 60 58 51 45 79 68 55 59 46 21 83 69 64 61 54 28 Mean religiosity, 100-point scale 73 54 53 58 Believe: V191. "Do you believe in God?" (%Yes) Life: V192. "Do you believe in life after death?" (% Yes) Importance: V196. "How important is God in your life." (% "Very," scaled 6-10) Identify: V186. "Independently of whether you go to church or not, would you say you are... a religious person, not a religious person, or a convinced atheist?" (% Religious) Comfort: V.197. "Doyou find that you get comfort and strength from religion?" (%Yes) Attend: V185. "Apart from weddings, funerals and christenings, about how often do you attend religious services these days?" (% Once a week or more) Source: WVS/EVS, pooled sample 1981-2001. 71�437�173053669560019791967 56 I. The Causes of the Rising Tide societies are clearly the most religious: in these nations, almost half of the public regularly attend church, while three-quarters or more regard themselves as religious, see religion as very important in their lives, and believe in life after death; and there is an almost universal belief in God. By contrast, postindustrial societies prove to be far less religious across all the different indicators, but especially on the behavioral item of church attendance, where only a fifth remain regular churchgoers. The industrialized societies prove to be moderately religious, but far closer to the rich than to poor nations. To examine the cross-national pattern more systematically, and to identify any important outliers, the scatter plot in Figure 3.2 compares strength of religiosity to indicators of societal modernization, measured by the country's rank on the standard UNDP Human Development Index, incorporating life expectancy, education and adult literacy, and GDP per capita. The results confirm that modern societies tend to be more secular in orientation. Simple economic development provides part of the explanation here, but an even stronger relationship is evident with human development; societies with widespread literacy, education, affluence, security, and access to multiple sources of information from the mass media tend to be the most secular. By contrast, the poorest and least developed nations, such as Bangladesh, Zimbabwe, Nigeria, Ghana, and India are the most religious in their values, beliefs, and behavior. The wide scatter of countries in the middle of the graph shows the continued impact of historical legacies of Orthodox faith in the post-Communist world in Central and Eastern Europe compared to the stronger role of the church in Latin America, where Catholicism has traditionally predominated. The comparison also confirms that the United States and Ireland are outliers on both measures, along with Italy, as countries that remain far more religious than most other comparable affluent postindustrial societies, with Scandinavian societies with a Lutheran background proving to be the most secular. China also remains a striking outlier, a pattern that may reflect both the confucian tradition and communist constraints on the open expression of religious feelings in that country. Moreover, as before, we can break down these patterns by cohort of birth and gender to see whether it is the younger generations in postindustrial societies who prove to be least religious, and whether cohorts of women differ more than cohorts of men, as indirect evidence Religion, Secularization, and Gender Equality 57 Malta Phil Zim Tag A Col jofrarf**1 ^ « Bng n ■ Ire Pol ABraz Mex & „ DomW-Sf " ■IS m A Tuffí India S Ar^1 ' asm ItSrepge ^ Geo B ^^0*-"*0*'^ B Can m 1 * SloukCro A.^ Alb Mol A JUth__________ Ar§i a AusSp «^ Mac ■ ■ Ice Sv& m ^■"Tjru * Ukr Jffetg^t-ux ST * ■swe|?" ľ Hun9 Lat A Bela^ Tan; m Est Sjjífcti ■* Suis Viet m China ■ Society * Agrarian ^ Industrial ^ Postindustrial Rsq = 0.1960 Human development (HDI Rank, 1998 High to Low) figure 3.2. Religiosity and human development. Note: For the 100-point Strength of Religiosity scale, see Table 3.1. Source: The rank on the Human Development Index, 1998, including measures of longevity, adult literacy and education, and per capita GDP, is from the UNDR Human Development Report, 2001. New York: UNDP/Oxford University Press. of long-term culture shifts. The graph in Figure 3.3 vividly confirms these propositions. In affluent nations, there is a steady fall in religious adherence among successive cohorts of women and men, so that the youngest generation is about fifteen points less religious on this scale than the interwar generation. But it is women in these affluent societies who have been transformed most by this process: the oldest cohort of women in postindustrial societies is more religious than similar cohorts in industrialized nations, yet the younger generation of women in postindustrial nations is by far the least religious. This suggests that the transformation in women's lives in modern societies during the twentieth century, generated by widening opportunities in education, the workforce, and public affairs and changes in families, the home, and modern lifestyles, has contributed to this dramatic decline in religiosity, along with broader trends in societal modernization 5« I. The Causes of the Rising Tide Women Agrarian ,4iii»»»»,!4 en 70 g .2> 60 o .n 50 en 40 Industrial Postindustrial ">> "•?> *&> % % ^ \ x x x Men I 50 o> c S čň 40 Agrarian .*»»** Industrial Postindustrial figure 3.3. Strength of religiosity by cohort, gender, and type of society. Note: The combined 6-item Strength of Religiosity Scale, standardized to 100 points, consists of the following items: Identify: Vi 86. "Independently ofwhetheryougo to church or not, would you say you are... a religious person, not a religious person, or a convinced atheist?" (% Religious) Attend: V185. "Apart from weddings, funerals and christenings, about how often do you attend religious services these days?" (% Once a week or more) Importance: V196. "How important is God in your life?" (% Very, scaled 6-10) Believe: V191. "Do you believe in God?" (% Yes) Life: V192.. "Doyou believe in life after death?" (% Yes) Comfort: V. 197. "Do you find that you get comfort and strength from religion?" (% Yes) Source: Pooled WVS/EVS, 1995-2001. Religion, Secularization, and Gender Equality 59 that have affected both sexes equally. Interestingly, industrialized nations show a modest slide in religiosity among the interwar generations of women, before a fairly stable pattern emerges among postwar cohorts. Lastly, in agrarian societies, aside from an early "blip" that is probably due to the limited number of respondents in the 1900-1916 cohort, there is a steady high level of religiosity. Many observers, such as Samuel Huntington, suggest that there has been a religious revival in many poorer societies in recent decades, especially among fundamentalist sects in Muslim societies, perhaps as a backlash against the perceived threats of modern Western values to traditional social norms and sexual mores.18 The cohort analysis suggests that agrarian societies remain largely unchanged, so that the younger generations are as fully religious as their parents and grandparents but not more so. But this results in a widening cultural gap, as affluent Western nations have become progressively more secular in orientation. Although some scholars of religions have disputed growing secularization, Weber's thesis - that modernization leads to less religious societies - is supported by analysis of age differences and by cross-national comparison of the survey evidence, although it is clear that the different historical legacies of different faiths continue to shape their worldviews. The churches, sects, and faiths founded centuries ago have left an indelible imprint on each society that remains evident in religious practices, beliefs about God, and patterns of religious attendance around the globe. The Values of Sexual Liberalization Just as societal modernization has affected beliefs in God, churchgoing habits, and the authority of religious leaders, so we would expect parallel developments to be evident in support for traditional moral values governing sexuality. We will examine four issues that have been an important part of the feminist debate, the women's movement, and sexual liberalization: approval or disapproval of abortion, divorce, homosexuality, and prostitution. Abortion involves the core claim that women should have control over their reproductive rights, the issue most hotly debated between liberals and the Christian Right in the United States, although one less contentious today in many European Men On O Postindustrial ^ ^~"~~ ^ „, ■.«»■ - « *** *« : o E o 3 -2 -1 * Industrial _, «> ** *" '"'" „,t ,„ «» '•** ** Agrarian figure 3.4. Sexual values by cohort, type of society, and gender. Note: Q: "Please tell me for each of the following statements whether you think it can always he justified, never be justified, or something in between, using this card from 1 (never justifiable) to 10 (always justifiable)." Homosexuality, prostitution, abortion, divorce. Source: WVS/EVS, 1995-2001. Women 7.0 6.0 5.0 4.0 c 0 3 3.0 Postindustrial --■''' s« ^ *iK *** tn O Q- 2.0 ■ 5 1.0 — Industrial»" * ,*»•" Agrarian )16 1927-1936 1947-1956 1967-1976 1917-1926 1937-1946 1957-1966 1977-1984 Men Postindustrial Industrial í «K «« ^' ,„........••'" Agrarian iiiii""" 1900-1916 1927-1936 1947-1956 1967-1976 1917-1926 1937-1946 1957-1966 1977-1984 Women Men Postindustrial Industrial iania>iaiiigia,at Agrarian 1900-1916 1927-1936 1947-1956 1967-1976 1917-1926 1937-1946 1957-1966 1977-1 Postindustrial «* Industrial Itlllllll1".....tllMIIIII Agrarian 1900-1916 1927-1936 1947-1956 1967-1976 1917-1926 1937-1946 1957-1966 1977-191 FIGURE 3.4 (coitt.). 62 I. The Causes of the Rising Tide societies. Divorce is an important issue, especially in a few Catholic and many Muslim societies, where women have limited or no legal rights to dissolve the marriage. Tolerance of homosexuality, including gay and lesbian rights, has also been a central part of the sexual liberalization that began during the early 1960s. And the issue of prostitution is one that divides the women's movement, although liberals have argued for greater sexual tolerance, and indeed for legalization and unionization to protect the rights of sex workers. If traditional religious values have eroded on a consistent basis, along with patterns of church attendance, then this should become clear through cohort analysis of approval of these forms of sexual liberalization, divided by type of society and by gender to see whether women and men agree on these values. On each issue, people were asked to use ten-point scales to show how far they believed abortion, prostitution, homosexuality, and divorce were justified, ranging from low (never justified) to high (always justified). Figure 3.4 shows the dramatic changes in sexual values evident among the older and younger cohorts in postindustrial societies, with steadily increasing tolerance for nontraditional sexual values across all four dimensions (with the exception of a modest conservative reversal among the youngest cohort on the issue of prostitution). The liberalization of sexuality in modern nations is particularly dramatic on the issues of abortion, divorce, and homosexuality, affecting the attitudes of both women and men, although women are slightly more tolerant of homosexuality and divorce. The pattern in industrialized nations shows broadly parallel shifts, although at a lower level of support, and there remain some important differences in the trends, for example, unlike the trend in more affluent countries, there is no "backlash" yet on tolerance of prostitution in these societies. By contrast, traditional sexual values are evident in agrarian societies, and there is usually little difference between younger and older cohorts, with the interesting exception of abortion, where there is evidence of a shift toward greater disapproval among younger generations. The overall pattern shows that the decline of religiosity we have found in postindustrial societies is not simply confined to replacing of spiritual beliefs and reliance upon religious authorities with a more rational and secular orientation, but that this change in the cultural outlook is closely associated Religion, Secularization, and Gender Equality 63 with major shifts in core sexual and moral values. The consistency of the generational differences across various indicators in each type of society strongly suggests that these shifts are not simply ephemeral phenomenon, responding to temporary fashions and fads, but instead represent enduring developments, so that younger women and men in affluent nations differ in many basic values from their parents' and grandparents' generations. Despite the common observation that the sexual revolution started in 1964, the patterns of moral values that became evident then are part of a far longer-term development with a momentum that continues to move younger generations in a more liberal direction. The Impact of Religiosity and Denomination on Gender Equality Does strength of religiosity and type of religious faith help to predict beliefs about sex roles as well as about gender equality in education, the workforce, and the public sphere - the issues at the heart of this study? If so, it suggests that increasing secularization, which, as we have seen, is consistently associated with societal modernization, may play an important part in explaining the rising tide of gender equality. Most work on this issue has compared the impact of religious cultures by classifying societies based on the predominant faith in each nation, drawing on common worldwide reference sources and almanacs. For the global picture, we can compare 190 nations based on the CIA's World Factbook zooi, with data supplemented by the estimates of religious populations drawn from the World Christian Encyclopedia. Figure 3.5 maps the classification of predominant religions in each nation based on these sources. Overall, nation-states divide into three major blocs: the largest number, fifty-seven nations, are classified as Roman Catholic, while forty-nine are Muslim and forty-six are Protestant, with the remainder Orthodox (twelve), Buddhist (twelve), other (including indigenous religions), Hindu, and Jewish. In terms of populations living under different majority religions, however, there is a different picture, as the largest group, representing over one billion people, live in Muslim states, and almost as many live in Hindu states (mainly India). Further analysis of how these denominations divide by type of society, shown in Table 3.3, demonstrates the strong O) m i ^ S O O U S o c o '5> U a o E/2 O E o m- ep O (8 c '0 •o * 3 £ a o o "? N nj O S ° D 'S u, a, 64 Religion, Secularization, and Gender Equality 65 table 3.3. Major religion by type of society, 190 nation-states worldwide Type of Society Major Religion Agrarian Industrial Postindustrial Total Roman Catholic T-9 2.9 9 57 Muslim 39 10 49 Protestant 28 9 10 47 Orthodox 2 10 12 Buddhist 8 3 1 12 Other 9 9 Hindu 2 1 3 Jewish 1 1 Total number of nation-states 107 63 20 190 Classification of types of society: See Appendix B. Classification of major religion (adhered to by the largest population) in 190 nation-states around the world based on CIA. The World Factbook, zooi. Washington, DC: Central Intelligence Agency. "Other" includes "indigenous beliefs." Source: relationship between modernization and Christianity (including Roman Catholicism and Protestantism), with all of the postindustrial societies around the world drawing on a Christian heritage, with the single exception of Japan. Nevertheless, it is apparent that there is a wide variety of sects and faiths among industrialized and agrarian societies, including Catholic, Protestant, Muslim, and Buddhist, as well as indigenous local religions. One important limitation of these classifications is that they take no account of pluralistic societies that are deeply and fairly evenly divided into many minority religious faiths, none of which includes a third or more of the population. There are important minority groups in societies such as Israel, Northern Ireland, and Bosnia-Herzegovina that are deeply divided by ethno-religious conflict. The categorization of "predominant religion" often measures nominal adherence to "official" denominations, reflecting birth certificates, practices of baptism, and census records, rather than the proportion of active practitioners or true believers. Moreover, other than estimates of regular religious attendance, based on the annual "mega-survey" published in the World Christian Encyclopedia, there are few reliable measures of the strength of adherence to religious beliefs and practices. The global comparisons of the impact on gender equality of the predominant religion in 66 I. The Causes of the Rising Tide table 3.4. Mean scores on the Gender Equality Scale by denomination, religiosity, and type of society Agrarian Industrial Postindustrial To Denomination No denomination 67 69 So 72 Catholic 69 70 81 71 Protestant 64 66 81 71 Jewish 72 64 85 67 Orthodox 64 62. 77 63 Hindu 62 74 70 6? Buddhist 70 60 61 61 Muslim 57 59 76 Í7 Religiosity Weak 63 67 81 70 Moderate 63 68 79 68 Strong 61 67 77 65 All 60 68 81 67 Note: For the coding, see Appendix B. For the 100-point Gender Equality Scale, see Table 2.1. The 100-point Strength of Religiosity Scale in Table 3.1 is recoded to produce the categories. Source: Pooled WVS/EVS, 1995-2001. 190 nations therefore need to be supplemented by more detailed analysis of the wide range of religious beliefs and practices in over 70 nations included in the pooled 1981-2001 World Values Surveys / European Values Surveys. Since we assume that basic religious values such as faith in God, attendance at church services, and denominational identities evolve only at a glacial pace, in response to long-term social and economic developments, the comparison of all societies across the pooled dataset is appropriate for cross-national comparisons. In order to analyze the impact of religion, Table 3.4 first describes the mean scores on the Gender Equality Scale used in the previous chapter broken down by adherence to different denominations and by strength of religiosity, without any prior controls. But for a more systematic understanding, we need to turn to multivariate regression analysis models, since we have already established that levels of human development and many social factors such as age, education, and gender affect both religiosity and support for gender equality. Table 3.5 provides the results of OLS regression models for the impact of denomination and Religion, Secularization, and Gender Equality 67 table 3.5. Religion and support for the Gender Equality Scale Model 1 Model 2 Unstandardized Standardized Correlation Coefficients Std. Beta (R) (B) Error Coefficients Sig. (Constant) 31.519 .782 .000 Religious denomination Catholics •13 7-75 .307 .194 .000 Protestants .07 4-43 .309 .110 .000 No denomination .10 4.41 •337 .090 .000 Hindu -.04 13.2 2.76 .018 .000 Jewish —.01 6.14 1.36 .017 .000 Orthodox (e.g. Greek —.10 0.66 •349 .012 .057 or Russian) Buddhist -.04 -4-47 .688 —.026 .000 Muslim —.20 -1.41 .626 —.010 .024 Strength of religiosity Religiosity -.19 -.008 .004 .Oil .062 Adjusted R2 .24 Model 1: Simple correlations without any prior controls; all results are significant at the .01 level. Model 2: OLS regression models with the Gender Equality Scale, 1995-2000 (five-item) as the dependent variable, controlling for societal modernization (Human Development Index, 1998; level of democratization, 1999) and social background (age, gender, education, income, frequency of religious attendance, work status, marital status, children, and family savings). For details of the full model and the control variables, see Table 2.3 and Appendix B. The religious denominations are coded as dummy (0/1) variables. Source: Pooled WVS/EVS, 1995-2001. strength of religiosity on the 100-point Gender Equality Scale 1995-2000, as the dependent variable, controlling for societal modernization (the 1998 Human Development Index and the 1999 Freedom House level of democratization) as well as social background variables (age, gender, education, income, frequency of religious attendance, work status, marital status, children, and family savings), factors all found to be important when used earlier in Chapter 2. Given the large size of the sample, the usual tests of statistical significance are not helpful in interpreting the results of the analysis, so we focus on the strength of the standardized beta coefficients. The results of this analysis demonstrate that the type of religion matters for beliefs about gender equality far more than the strength 68 I. The Causes of the Rising Tide of religiosity. In Table 3.4, the strongest contrast in attitudes toward the appropriate division of sex roles among women and men is that between Western Christian and nondenominational populations living in affluent postindustrial societies, who adhere to the most egalitarian beliefs about the family, workforce, and politics, and Muslims living in poorer, agrarian nations, who are by far the most traditional group in their attitudes toward gender equality. Moreover, these patterns continue to prove significant in the multivariate analysis presented in Table 3.5, even after controlling for levels of societal and democratic development and the standard social variables such as age, gender, and education that we have already established as explanatory factors. This suggests that religious beliefs are not simply a by-product of the fact that many Muslims can be found living in poorer societies in North Africa, the Middle East, and Asia; instead, the evidence indicates that traditional religious values and religious laws have played an important role in reinforcing social norms of a separate and subordinate role for women as homemakers and mothers, and a role for men as patriarchs within the family and primary breadwinners in the paid workforce.19 If we compare attitudes toward both gender equality and the indicators of sexual liberalization discussed earlier (including approval or disapproval of divorce, abortion, prostitution, and homosexuality) by predominant type of religion, rather than by type of socioeconomic development, the results confirm two striking and important patterns. First, there is a persistent gap in support for gender equality and sexual liberalization between the West (which is most liberal), Islamic societies (which are most traditional), and all other societies (which fall between these extremes). Another finding has very important implications: the gap between the West and Islam is usually narrowest among the oldest generation, but this gap has steadily widened across all the indicators as the younger generations in Western societies have become progressively more liberal and egalitarian, while the younger generations in Islamic societies remain as traditional as their parents and grandparents. This suggests that Islamic societies have not experienced a backlash against liberal Western sexual mores among the younger generations, as some popular accounts assume, but rather that young Muslims remain unchanged despite the transformation of lifestyles and beliefs experienced among their peers living in postindustrial societies.20 But the result is Religion, Secularization, and Gender Equality 69 a growing gap between the basic values of young people in Islamic societies and those in the West. Moreover, these predominant beliefs and values matter not only for cultural attitudes, but also for the actual conditions of men and women's lives. Table 3.6 shows the relationship at the national level between the proportion of the population adhering to different religious denominations, the strength of religiosity, and a range of common social indicators of gender equality selected from the UNDP Human Development Report 2001 as the dependent variables. The cross-national comparison is restricted to the ninty-three poorer, agrarian societies around the world, in order to isolate the impact of religious denominations by controlling for countries at similarly low levels of human development. The differences in the indicators cannot be explained simply as by-products of low per capita income or the restricted availability of educational opportunities or maternal health care, since all of the nations under comparison are classified with a Human Development Index of .739 or below and a per capita annual income of about $1,000 or less. The male/female ratio measures that are used also provide the fairest indicators, as this statistic focuses on relative disparities between women's and men's lives within each of these societies, rather than on absolute levels of education or development. The majority religion in each nation is classified based on the CIA World Factbook, zooi. The results of the analysis, shown in Table 3.6, confirm that religion matters, not only for cultural attitudes but also for the opportunities and constraints on women's lives, such as the ratio of females to males in educational enrolment, the female adult literacy rate, the use of contraception, and the UNDP Gender-Related Development Index, as well as for opportunities for women in the paid workforce and in parliamentary representation. Conclusions and Discussion An extensive literature in sociology, anthropology, social psychology, divinity studies, and women's studies has long argued that religion has exerted a decisive impact on the cultural perceptions of the appropriate division of labor between men and women, and that it has shaped social norms and sexual values. The influence of religious authorities s C/l 3 «5 O 00 NO i/-i H m H OO < H (n| vo vo oo |\ 1^1 ^3 3 « u "~i 1/-5 on a,7j g ft ."> Já e « e S c 1> o tu « o 2 Ö 'S ü E "S -o a ti ■- O H 00 00 h- i/-, ON ON ON 00 00 N 0\M N ti "J 4-T |J S J "g J ."P e 75 s .a -S « ■i 3 .-o 3 S -3 I S -o c o —' >-« if y ^ C cu JU T3 "ŕŕí C! 3 H I/I 1/-5 m 1/1 r) VO M 00 00 K 1/-1 00 On l/N ON ON ■f 00 ON i H |\ ON >/~, NO l/"5 ■f 1^ ON ON ON H H O 0 On On M On On H On On H ON ON ON ON H 0 ON ON H 1 M ON On H On ON H On ON H -2 u n3 3 —C > Q -f — w "c « ľ! =3 c S S o s g « e* ~ O O g CD C/5 r, E w P3 OJ Oh (Ä Ü « D Cj (U (U JJ .> o B- ^ — Q "^ ,/ V U U W E § U* £5 "rt "« O 0-( E S "oj .E ~ct> .5 « t3 t! « 55 11 t* r ^ % u 2 c I- ŕ (U w — II .2 | S w C iL ■*!"° "£ "° 3, c .SO« tí >- C5 Pt CU W5 U ^ flj tcQ- 'S2 >. iS c 2 'S O 2 S SP u « _ß CU *-* I ■a § o A J* O o -Q y; o C 5) .2 »í C CÄ 'S B —, ra os C/5 c/5 C cu .b: -o — _3 c "u 2 c a •— « c -= S ä S H. c s E w -G cu H -H .y s g 3 S 3 3 §• c di -S —' > M u O S -C 00 « *-" « S Sí> u á -S í*1S x: q .a v 2 -a c\j ID • — Ses s o s ci y i-v 7o Religion, Secularization, and Gender Equality 71 has often served to limit opportunities for women outside the home, in education, at work, and in positions of authority. The results of this study confirm that religion plays an important role in this regard; in particular, the study establishes four key findings: 1. A process of secularization has gradually accompanied societal modernization, weakening the strength of religious values among the younger generation in postindustrial societies and fuelling the rising tide of gender equality. 2. Cohort analysis shows that postindustrial societies have experienced a parallel liberalization of moral values regarding sexuality among the younger generation, exemplified by attitudes toward the issues of abortion, homosexuality, prostitution, and divorce. 3. At the same time, religiosity continues to exert a strong influence on social norms about the appropriate division of sex roles in the home, the workforce, and in the public sphere, especially in agrarian societies. 4. Attitudes towards women vary among adherents of different religious sects and denominations; in particular, an Islamic religious heritage is one of the most powerful barriers to the rising tide of gender equality. It remains to be seen how far the coherent cultural shifts in attitudes toward gender equality that we have uncovered matter for gender differences in public life, particularly among the electorate, in civic society, and in political leadership. It is to these matters that we now turn. PART II THE CONSEQUENCES OF THE RISING TIDE There is a tide in the affairs of men Which, taken at the flood, leads on to fortune. Shakespeare, Julius Caesar 4 The Gender Gap in Voting and Public Opinion We have explored the causes of the rising tide of gender equality, but not yet the consequences. During the postwar era, the conventional wisdom in political science held that women in Western democracies were politically more conservative than men. Gender differences in party preferences were never as marked as the classic electoral cleavages of class, region, and religion; there were no mass "women's parties" like those associated with trade unions, regions, and churches. Nevertheless, "women's conservatism" was seen as a persistent and well-established phenomenon. During the 1980s, this conventional wisdom came under increasing challenge. In many West European countries, a process of gender dealignment appeared, with studies reporting minimal sex differences in voting choice and party preference. And in the United States, the phenomenon of the gender gap manifested itself in the early 1980s, with women shifting their allegiance toward the Democratic Party while men moved toward the Republican Party on a stable and consistent basis, reversing the previous pattern of voting and partisanship.1 This gender realignment in the United States raises the question of whether similar developments are occurring elsewhere. If this phenomenon is caused by factors inherent in societal modernization, such as increased female participation in the paid workforce, the break-up of the traditional family, and the transformation of sex roles in the home, then we would expect to find similar gender gaps emerging in other postindustrial nations. On the other hand, if the gender gap is 75 76 II. The Consequences of the Rising Tide caused by factors that are distinctive to American politics, such as its lack of strong class cleavages, the centrist pattern of two-party competition, and the salience of issues such as abortion and affirmative action, then the modern gender gap in the United States might be still another example of American exceptionalism.2 Answering this question is important in order to clarify whether the rising tide of gender equality is altering women's political power in many nations, and also to provide a clearer theoretical understanding of the underlying process of value change. This chapter focuses on comparing gender differences in ideology, electoral preferences, and public opinion, the most common meaning of the term "gender gap," in more than seventy nations. Building on previous chapters, we hypothesize that (i) the process of societal modernization is reshaping the political values and attitudes of women and men, just as it has altered other basic values. (2) In particular, women are now moving to the left of men, even controlling for structural differences in men's and women's lives such as religiosity, education, and workforce participation. (3) Because this transition is still taking place, we would expect to find substantial variations in this pattern by age cohort in postindustrial societies, with older women remaining more conservative than men, while the younger generation of women move to the left of men. If this is indeed the case, these societies would seem to be experiencing an intergenerational ideological realignment in gender politics that is likely to persist, given the long-term process of generational replacement. Theoretical Framework The Orthodox Account of Female Conservatism Research on gender differences in the electorate has been a recurrent theme in political science beginning with the earliest systematic surveys of voting behavior.3 Many hoped, and others feared, that once women were enfranchised there would be a distinctive "women's vote." Gender was not regarded as a primary electoral cleavage, equivalent to class, region, and religion, because women and men experienced many cross-cutting forces, but the seminal account of European voting behavior by Lipset and Rokkan viewed gender as one of the secondary cleavages shaping the electoral base of party politics.4 The Gender Gap in Voting and Public Opinion jj table 4.1. The gender gap in the early iyjos Men Women Gap Italy 44 30 -14 Germany 60 47 -13 Britain 50 41 -9 Belgium 40 36 -6 France 54 49 -5 The Netherlands 47 45 —2 United States 32- 37 +5 Note: Percentage supporting parties of the left. Source: Ronald Inglehart. 1977. The Silent Revolution: Changing Values and Political Styles among Western Publics. Princeton, NJ: Princeton University Press. The early classics of the 1950s and 1960s established the orthodoxy in political science: gender differences in voting were generally fairly modest, but women were likelier than men to support center-right parties in Western Europe and in the United States, a pattern that we can term the "traditional gender gap."5 Most explanations of the traditional gender gap emphasized structural differences between men and women in religiosity, longevity, and labor force participation; for example, women in Italy and France were more likely to attend churches associated with Christian Democratic parties.6 During this era, women were also commonly assumed to be more conservative in their political attitudes and values, producing an ideological gap underpinning their party preferences.7 The conventional wisdom was summarized in The Civic Culture, first published in 1963: "Wherever the consequences of women's suffrage have been studied, it would appear that women differ from men in their political behavior only in being somewhat more frequently apathetic, parochial, [and] conservative___Our data, on the whole, confirm the findings reported in the literature."8 Inglehart confirmed that during the early 1970s, women remained more likely than men to support Christian Democratic and Conservative parties in Western Europe, particularly in Italy and Germany (see Table 4.1), although a new pattern seemed to be emerging in the United States, which he speculated might represent the wave of the future.9 Nevertheless, the prevailing view was that the relative conservatism of women was a fixed, structural characteristic. 78 II. The Consequences of the Rising Tide Theories of Gender realignment This orthodoxy came under increasing challenge during the 1980s, when scholars in many Western countries found a pattern of gender dealignment among the electorate. Voting studies noted this pattern in Britain10 as well as in Germany,11 the Netherlands,12 and New Zealand.13 This literature suggested that the old pattern of female conservatism was apparently no longer evident; instead, the situation in the 1980s seemed contingent upon political circumstances: in some established democracies women seemed to lean to the right, in others to the left (particularly in Nordic societies),14 and in others no significant differences could be detected.15 Studies of ideological self-placement, rather than voting choice, found that during the mid-1980s women in Western Europe saw themselves as slightly more conservative than men, although this gap was reduced when controls were incorporated for labor force participation and religiosity.16 The pattern seemed consistent with dealignment theories suggesting that the impact of traditional social-party linkages had weakened in many established democracies, notably the force of social class and religion.17 These theories argued that voters had become more instrumental. Under these conditions, no party could expect to enjoy a persistent and habitual advantage among women or men; instead, contingent factors such as government performance, party policies, and leadership images would dominate voting decisions. Theories of Gender Realignment During the last decade, however, there has been much speculation, although little concrete evidence, that women were realigning to the left throughout postindustrial societies, a situation that we will term the "modern gender gap," replicating the pattern that first emerged in the United States. The process of "partisan realignment" is understood to produce an enduring and stable change in the mass coalitional basis of party politics. A classic example occurred in the United States when African-Americans moved toward the Democrats during the 1950s and 1960s, while Southern white conservatives shifted toward the Republicans, leading to a persistent and deep-rooted change in the racial basis of American party politics.18 In the United States, the process of gender realignment meant that although women leaned toward the Republican Party in the 1952,1956, and i960 presidential elections, The Gender Gap in Voting and Public Opinion 79 during the 1960s and 1970s traditional gender differences in the electorate faded, and from the 1980s onward the modern voting gap became apparent in successive presidential, gubernatorial, and state-level contests, as well as in levels of Democratic Party identification.19 The modern gender gap in American elections has not been as large as cleavages based on race or religion; nevertheless, this pattern has proved consistent, stable, and decisive in many contests, representing a long-term shift in the mass basis of electoral competition in the United States. The Developmental Theory of Gender Realignment What might cause gender realignment to occur? We hypothesize that structural developments lead to, and interact with, cultural shifts that tend to reshape political values. The long-term trajectory of value transformation is generally predictable and coherent, although the pace of realignment within each country remains contingent upon institutional contexts, the political mobilization of major social cleavages, and other situation-specific conditions. Factors affecting this process include how far leaders respond strategically to shifts in public opinion, which groups are mobilized into politics, and patterns of party competition. Previous chapters have demonstrated how preindustrial and agrarian societies continue to be characterized by sharply differentiated gender roles that discourage women from taking jobs outside the home. As we have already seen, virtually all preindustrial societies emphasize traditional sex roles: childbearing and child rearing are regarded as the central goal for women and their most important function in life; careers in the paid workforce outside the home remain predominately male. Religiosity reinforces traditional sex roles. By contrast, we have shown how gender roles converge in postindustrial societies due to the cultural shift in attitudes toward women and the structural revolution in the paid labor force, in educational opportunities for women, and in the characteristics of modern families. This revolution occurred during the twentieth century in the most developed nations, producing contrasting experiences for the younger generation of women compared to their mothers and grandmothers. As women's lives change in postindustrial societies, we expect that this process will influence broader social values, political attitudes, ideological orientations, and ultimately partisan preferences. 8o /. The Consequences of the Rising Tide This account emphasizes that structural and cultural trends common to postindustrial societies have led to secularization and to more egalitarian sex roles, both of which have realigned women's political values, particularly among the younger generation. These hypotheses generate certain testable propositions about how the size and direction of the differences between women and men in left-right political ideology and in left-right voting behavior should vary systematically, under four conditions: i. First, if we analyze trends over time, we would expect to find that women in recent decades have gradually shifted to the left in their voting behavior and political ideology. 2. The transformation of sex roles in the paid labor force, in education, and in the family has gone much further in postindustrial than in preindustrial societies, as has the process of value change. Consequently, in cross-national comparisons, we would expect to find that the modern gender gap has advanced furthest in postindustrial societies. 3. If we compare generational cohorts within postindustrial societies, as indirect evidence of longitudinal change, we would also predict that the ideological gender gap would reverse by birth cohorts, given the way that changes in lifestyles and cultural trends have transformed the lives of older and younger groups of women in these nations. Previous studies in Britain have examined the pattern of the gender gap by generation, with younger women leaning left, while older women remain more conservative than their male counterparts.10 Consequently, we would hypothesize that in these countries, the modern gender gap should be evident among the younger generation, and the traditional gap should remain relatively strong among the older cohorts. These generational patterns would not be expected in industrialized or developing societies, since structural and cultural changes in these societies have taken different pathways. 4. Finally, if we compare social groups within societies, we would expect to find that women today hold more left-oriented values than men, in terms of support for an active role for government in social protection and public ownership of business and industry, and in support for a variety of political issues. Support for The Gender Gap in Voting and Public Opinion 81 left-leaning values among women and men can be expected to vary systematically according to structural factors, namely, participation in the paid labor force, class, education, marital status, union membership, and religiosity, as well as according to cultural factors, including attitudes toward gender equality, general beliefs about the role of government, and postmaterialist values. Ideological Values and Voting Behavior Attitudes toward the Role of Government Does the evidence support these propositions? One can compare left-right economic values using a variety of indicators, but perhaps the most common measure suitable for cross-national analysis concerns attitudes toward the role of the state and the marketplace. The classic left-right issue cleavage, long dividing socialists and conservatives, concerns how far government should go in providing social protection and an economic safety net via the welfare state, and how far the private sector and the free market should be unconstrained by the state. The World Values Surveys / European Values Surveys, 1990-2000, tap these values with two items that form a single reliable indicator; here they are recoded so that high scores represent the conservative position and standardized to a 100-point Role of Government Scale, for ease of interpretation. The items ask the following: "Now I would like you to tell me your views on various issues. How would you place your views on this scale? 1 means you agree completely with the statement on the left. 10 means you agree completely with the statement on the right, and if your views fall in-between you can choose any number in-between. * Private ownership of business and industry should be increased (10) - Government ownership of business ands industry should be increased (1). • The government should take more responsibility to ensure that everyone is provided for (1) - People should take more responsibilities to provide for themselves (10)." Comparing the mean scores on the Role of Government Scale, as shown in Figure 4.1, confirms the expectation that across most countries of the world, women today hold more left-leaning views than men, with the gender gap being particularly strong in poorer nations such as Uganda, Ghana, and Bangladesh, as well as in some middle-income 82 II. The Consequences of the Rising Tide ~Ugaří3f~É| Ghai^a I Latvia, Dominiqan Rep Šíovenia L Bangladesh Moldova I Japan i Armenia f Georgia ■ Estonia I Norway 1 Belarus 1 Hungary I Romania I Portugal B Lithuania I Serbia! I Russia I Taiwan | Finland I Germany 1 Zimbabwe L Poland L Germany | Czech Republic f Canada L Bosnia L El Salvador 1 Italy! Peru I Bulgaria ft Croatia 1 Netherlands L. Denmark L i Sweden ft : Iceland L : Slovakia | United States l Nlreland L Macedonia L Indian. Ireland I Azerbaijan HH Ctiile i^^m China wmmmm Argenfina MM Belgium BBBM Mexico BBHIM Venezuela nnni France MB United Kingdom M Iran HW Switzerland Hi ! Spain ftH 1 Turkey HB 1 Austria w 'Egypt H iAIbania B New Zealand n i Nigeria fti i Jordan | I Philippines I South Afpica M Uruguay BMMj Australia mm™ Korea, Rep.qf BBBBnnBLVietnami •12.00 -10.00 -8.00 -6.00 -4.00 -2.00 0.00 2.00 4.00 6.00 Women more left-wing than men - women more right-wing than men figure 4.1. Role of Government Scale, 1990-2.001. industrializing nations such as Latvia and Brazil, and in some postin-dustrial nations such as Japan and Norway. In some countries, however, the gender gap is minimal, and there are even a few nations where women are more conservative than men in their views of the role of government, such as Vietnam and South Korea. On this indicator, The Gender Gap in Voting and Public Opinion 83 at least, the evidence supports the thesis that women are overwhelmingly more favorable to an active role for the state, as other studies of American public opinion have found.11 Ideological Values But have the broad ideological positions of women and men changed? The longest cross-national series of surveys monitoring ideological positions is the Eurobarometer studies, conducted annually in the European Union countries from the early 1970s to the present. All of the member countries have affluent postindustrial economies, so this does not provide comparisons across rich and poor societies; but the series does provide the best cross-national evidence of long-term trends in how women and men have seen themselves ideologically during the last thirty years. The Eurobarometer surveys ask people to place themselves on a ten-point scale ranging from left (low) to right (high). This measure of political ideology is particularly appropriate for cross-national research, since there is evidence that most people are familiar with the idea of a left-right scale and that it has greater validity in multiparty contexts, at least in Western Europe, than the directional component of party identification.12 The results of the comparison, shown in Figure 4.2, confirm that the position of European men has remained relatively stable since the early 1970s, with some trendless fluctuations around the mean: in 1973, European men placed themselves at 5.25 on the left-right scale, compared to an average of 5.29 in 1999. In the early 1970s, European women placed themselves ideologically to the right of men, but the gender gap had closed by the mid-1980s, and by the mid-1990s women had shifted to the left of men. In 1973, women on average placed themselves at 5.61 (slightly to the right of men); thirty years later, they placed themselves at 5.20 (slightly to the left of men). Moreover, this pattern proved reliable in a broader comparison; similar trends were evident when we examined the eleven nations included in every wave of the WVS/EVS since 1981 (Britain, West Germany, Italy, Spain, the United States, Japan, South Africa, Hungary, Sweden, Argentina, and Finland). In the early 1980s, the women in these countries placed themselves to the right of men on this ideological scale; in the early 1990s, there was no significant difference; and by the mid-1990s, the position had reversed - women had moved to the left of men. 84 IL The Consequences of the Rising Tide - - Men ■ Women figure 4.2. Trends in the ideological gender gap, 1973-99, EU-5. Note: This shows the trends in left-right ideological self-placement among the five countries included in the surveys throughout this period (Italy, France, Germany, Belgium, and the Netherlands). Source: Eurobarometer surveys, European Union. Electoral Behavior Ideological values can affect electoral behavior, especially where party competition revolves around programmatic differences on economic issues such as the role of public ownership or the welfare state; but in many countries other issues may prove more important. Direct evidence of trends in the electoral gender gap can be compared in eleven postindustrial societies where information about voting preferences is available in successive waves of the WVS/EVS since the early 1980s. Respondents were asked: "If there were a national election tomorrow, for which party on this list would you vote?" The parties are classified along the left-right spectrum using the 1993 Huber and Inglehart scale based on expert ratings of the positions of parties in forty-two countries along a ten-point left-right scale.23 This measure allows comparison of the mean left-right positions of women and men voters within each country, calculating the voting gap as the difference between group means, tested for significance with ANOVA.14 The results in Table 4.2 show a mixed pattern for the early 1980s: in four countries (the Netherlands, the United States, Denmark, and Italy) women leaned to the left, while in six others they leaned rightward. But the pattern of change over time is consistent: in countries where women were more The Gender Gap in Voting and Public Opinion 85 table 4.2. Trends in the voting gap in the 1980s and 1990s 1981 Gap 1990 Gap 1995 Gap The Netherlands 08 51 ** Belgium — 39" 08 France — 39* 09 Canada 01 23 Britain — 2-5 03 West Germany — 06* 16* .05 Ireland — 28 20 United States 14 15" •35" Spain — 08 .21 * -.28 ** Denmark 84" .69 ** Italy 39** .05 Note: The difference between the mean position of women and men on the ten-point voting scale. A negative figure represents women more conservative than men; a positive figure represents women more left-leaning than men. Sig. ** p. = 01, * p. = 05. Source: WVS/EVS, 1981-95. conservative in 1981, this tendency weakens over time, although it does not disappear entirely, except in Spain. The modern gender gap, with women being more left-leaning than men, is evident in every wave in the United States; it consolidates over time in the Netherlands; and it emerges by the 1990s in Canada and West Germany.25 Cohort Analysis of Ideological Change Unfortunately, direct cross-national evidence that could be used to examine voting behavior and ideological trends over many decades in a wider range of societies is unavailable. As in previous chapters, however, we can utilize indirect evidence available through cohort analysis to compare the size and direction of the ideological gap among older and younger generations in over seventy nations included in the pooled WVS/EVS, 1981-2000. Differences between younger and older age groups may be the result of life-cycle effects; social and political values may change as people enter the workforce, marry and have children, or enter retirement. These effects cannot be ruled out on the basis of cross-sectional survey evidence. Cohort effects emerge when formative experiences during childhood, adolescence, and early adulthood leave an enduring imprint on basic social values and core political attitudes. 86 II. The Consequences of the Rising Tide Consequently, societies experiencing rapid societal modernization and human development should display contrasts between the older generations - those who grew up during World War II and earlier - and succeeding cohorts. In the WVS/EVS surveys, as in the Eurobarometer surveys, political ideology is measured by asking people to place themselves on a ten-point scale ranging from left (i) to right (10). This measure allows consistent comparison within a country at different periods of time, avoids the problems of coding voting preferences on a left-right scale for parties experiencing rapid change, such as during the democratic transition in Central and Eastern Europe, and avoids limitations on the reliability of expert judgments. We calculated an "ideological gap" by finding the difference in the mean self-placement of women and men on these scales; for example, if women place themselves at 5.5 and men at 6.8, the resulting ideological gap is +1.3. For consistency in the analysis, a negative figure indicates that women in a given country see themselves as more conservative than men (the traditional gender gap), while a positive figure indicates that women place themselves further left than men (the modern gender gap). Again, ANOVA is employed to test the statistical significance of the differences in the group means within each society. Figure 4.3 provides evidence suggesting that the gender gap in political ideology is consistently linked with the process of modernization: different patterns of cohort change emerge in different types of society. In postindustrial nations, the pre-war cohorts display the traditional gender gap, with women consistently placing themselves to the right of men. The ideological gap closes in the middle cohorts, and then reverses polarity among the postwar cohorts, so that younger women consistently place themselves to the left of younger men. The consistency of the slope from older to younger female cohorts suggests a gradual shift among women, who have steadily and persistently moved from right to left. By contrast, there is a fainter pattern in these societies among men, who have remained relatively stable across successive cohorts. Industrialized societies at lower levels of development show a more stable pattern among cohorts, although even here, in the older cohorts women place themselves slightly to the right of the men, while younger women place themselves slightly to the left of younger men. The preindustrial societies show no evidence of cohort change - but, contrary to expectations, women are consistently slightly to the left of The Gender Gap in Voting and Public Opinion Postindustrial 1897-1906 1917-1926 1937-19.16 1957-1966 1977-191 1907-1916 1927-1936 1947-1956 1967-1976 Cohort of birth 87 Industrial 6.5 - I 6.0 - Women 1 i 5.5 - 1 5.0 Men 1907-1916 1927-1936 1947-1956 1967-1976 1917-1926 1937-1946 1957-1966 1977-1984 Cohort of birth Agrarian 1917-1926 1937-1946 1957-1966 1977-191 1927-1936 1947-1956 1967-1976 Cohort of birth figure 4.3. Ideological left-right self-placement by gender and type of society. 88 IL The Consequences of the Rising Tide men. This suggests that theories of electoral competition that have been developed to explain partisan loyalties in the context of programmatic parties in Western Europe may not apply in developing nations, where parties are based more on personalistic appeals, leadership factions, and ethnic loyalties. The pattern of longitudinal ideological trends over thirty years in the Eurobarometer surveys, the patterns of voting behavior in the World Values Surveys / European Values Surveys beginning in the early 1980s, and indirect evidence from cohort analysis all point to the conclusion that the modernization process tends to bring realignment to the left among women, with the greatest evidence of ideological change being evident in the richest nations, where women have experienced the greatest transformation in their lifestyles and sex roles. No comparable realignment seems to have occurred in poorer societies. Explaining Gender Realignment To understand support for left-leaning values, and to examine the reasons for gender realignment in postindustrial nations, the roles of both structural and cultural factors need to be considered. Structural Explanations Structural accounts explain gender differences in electoral behavior, public opinion, and left-right ideology by women's and men's lifestyles and roles in the workforce, home, and family. In Lipset and Rdkkan's classic theory, social class, region, and religion were viewed as the most important political cleavages in many West European countries, because they reflected broadly based and long-standing social and economic divisions within society at the time when the Western democracies were emerging.16 Contemporary party systems resulted from complex historical processes, notably the national and industrial revolutions experienced by societies from the seventeenth century onwards. In Europe, the cleavage between church and state produced religious support for Christian Democrat parties; the cleavage between landowners and industrialists helped to create agrarian parties; and the cleavage between employers and workers generated the parties of the left. Thus groups based on social class, religion, language, ethnicity, and region became the primary building blocks for the political system. Parties The Gender Gap in Voting and Public Opinion 89 mobilized coalitions of social groups and appealed to their interests. By contrast, gender was usually regarded as secondary, since women's interests were seen as divided by crosscutting cleavages such as class, ethnicity, and generation. The varying pattern of social cleavages across Europe in the nineteenth and early twentieth centuries established the essential framework for contemporary party systems. After the systems were established, Lipset and Rokkan suggest, they "froze" as parties strengthened links with their supporters and integrated new social cleavages. In most countries, women won the franchise after the modern party system was established, and they were therefore absorbed into the existing framework. The gender gap can be explained within this general theoretical framework in two alternative ways. For dealignment theorists, the development of the modern gender gap is seen as part of a broader loosening of the traditional ties between social groups and parties, particularly the weakening of class alignments.27 The process of secularization that we have already examined can be expected to play a major role in cultural and political change. In the past, women tended to be more religious; Lipset argues that this, and particularly women's Catholicism, helped to explain greater female support for Christian Democrat parties during the postwar era.2,8 As already observed, trends in secularization have gradually weakened religiosity in affluent nations, eroding church-party linkages, although there is far greater religiosity in poorer preindustrial societies. Since the early 1970s, party fragmentation has grown in many established democracies, symbolized by the sudden rise of new regional, xenophobic, and Green parties in countries such as Italy, Canada, France, and Germany. New cleavages in society, produced by the rise of new value priorities, have changed the priorities of the policy agenda, with a decline in the old left-right politics of redistribution and rising emphasis on issues involving the environment, women's rights, and the quality of life/9 The old cleavages of class and religion have declined in importance, opening the way for the politics of gender, region, and ethnicity to become increasingly salient. A related explanation argues that structural change in postindustrial nations has produced a divergence in the socioeconomic position of women and men, so that gender can be seen as a basic social cleavage reflecting distinctive political interests. In this view, women's S>o 71 The Consequences of the Rising Tide new roles within the family, the labor market, the welfare state, and the community may be expected to lead to different patterns of political participation, partisan loyalties, and political priorities on such issues as child care, family support, public transport, the environment and technology, reproductive rights, welfare, education, and defense. If this process has affected the gender gap in voting behavior, considerable differences in the size of the voting gap could be expected depending on levels of female education, income, age, occupation, and marital status. But findings from previous studies in the United States are divided concerning this explanation. Miller concluded that the socioeconomic experiences of American women and men are insufficient to explain gender differences in the vote, since the gender gap persisted even after controlling for many demographic and social characteristics.30 But in another study, Susan Carroll noted that the gender gap was strongest among two groups: on the one hand, professional, college-educated, and fairly affluent groups, and on the other hand, the less well-off and unmarried groups.31 The common factor linking women in these two groups, Carroll suggested, is economic and psychological autonomy from men. A continuing pattern of gender differences by high- and low-income group has been confirmed by more recent American studies.3i Women's increasing entry into higher education can also be expected to influence their political values and priorities, as many studies of public opinion have reported that education is consistently associated with more liberal attitudes on a wide range of issues, including feminist attitudes.33 In a related argument, Klein suggests that the entry of more women into the paid labor force in postindustrial nations may have altered their objective economic interests, which may in turn encourage their support for parties of the left.34 Klein argues that women's work experiences fundamentally alter their roles and expectations, while the social context facilitates a social network with different values. As more and more women enter the paid labor force, they gain direct experience with pervasive patterns of horizontal and vertical occupational segregation.35 Working women are often overrepresented in low-paid jobs, experiencing pay disparities, lower socioeconomic status, and higher levels of female poverty.36 Moreover, mobilizing agencies have responded to the feminization of the workforce; in recent decades, trade unions have made increasing efforts to expand their membership base by recruiting working The Gender Gap in Voting and Public Opinion S>i women,37 a process that can also be expected to move this group towards the left. Lastly, marital status and the presence of dependent children have often been regarded as particularly important structural factors explaining the gender gap in America.38 Women who are divorced or do not marry have different lifestyles and interests from those who are married or cohabiting, and these differences can be expected to influence their perceptions of women's roles in the family and society and their attitudes toward broader issues such as family policy. In this view, therefore, the electoral behavior and political values of women and men should be predicted by the standard social and demographic background variables available in surveys, including religiosity, education, participation in the labor force, income and socioeconomic status, trade union membership, marital status, and children.39 Cultural Explanations Structural factors can be viewed as interacting with, and causing, shifts in cultural attitudes and values that may subsequently exert an independent effect upon voting choice. We have already demonstrated the contrasts in beliefs about egalitarian versus traditional sex roles; and the Gender Equality Scale used in Chapter 2 can be expected to predict broader political and ideological differences as well. Post-modernization theory suggests that the growth of postmaterialist values among the younger generation in postindustrial societies has led to a gradual erosion of class-based politics, opening the way for greater priority being given to the values of freedom, self-expression, and gender equality.40 This pervasive cultural shift has increased the salience of issues such as reproductive choice, sexual harassment in the workplace, and equal opportunities, although a time lag can be expected between the emergence of new issues on the policy agenda and the response of the party system.41 If the modernization process has influenced gender realignment, support for postmaterialist values should be closely associated with female support for parties of the left. In addition, Conover has argued that the electoral gap in America has been strongly influenced by feminism, the women's movement, and mobilization around issues of gender equality.42. In this view, the growth of feminist identity and consciousness has been the catalyst producing the modern gender 92 II. The Consequences of the Rising Tide gap in party support.43 If this is the case, we would expect to find the modern gender gap in postindustrial societies to be stronger among feminist women. Moreover, there may be broader attitudinal differences between men and women that give rise to gender differences in electoral behavior and partisan support. Studies of public opinion in the United States and Western Europe have found that women often display relatively strong support for government spending on the welfare state, public services, and environmental protection, as well as opting for pacifism in the deployment of military force.44 These latent gender differences in public opinion can be expected to become manifest in electoral preferences when these issues become salient and divide the major parties. The predominant ethos in Conservative and Christian Democratic parties of the right usually favors traditional roles for women in the home and family, equal opportunities rather than affirmative action to remove problems of sex discrimination in the marketplace, a minimal role for the state in social protection policies, and a stronger role for defense and security services. By contrast, reflecting their core beliefs and overall philosophy, parties of the left have been committed to maintaining a comprehensive welfare state and a strong social safety net for public services such as health, pensions, education, transportation, and child care, as well as pacifism in foreign policy. Women are often the prime beneficiaries of government services such as pensions and child care, as well as constituting many of the employees in professional, administrative, and service work in the health care and educational sectors. Although some Labour, Socialist, and Communist parties retain traditional orientations toward the division of sex roles, these parties usually promote more egalitarian policies toward women, such as the adoption of gender quotas in party posts and candidate recruitment for elected office.45 Analysis and Results Before analyzing the relative impact of structural and cultural explanations, we will first examine the mean conservative ideology scale among women and men in postindustrial societies, using the 1995-2002 World Values Surveys / European Values Surveys. Table 4.3 shows that the gender differences across most categories are modest, with most people table 4.3. The Left-Right Ideology Scale by gender in postindustrial societies, 1995-2001 Women Men Diff Education High 5.25 5-43 0.18 Moderate 5.48 5-56 0.08 Low 5.60 5.46 —0.14 Age group Under 30 5.12 5.26 0.14 30-59 5-37 5-39 0.02 60+ 5.82 5.72 —O.IO R's occupational class Manager/professional 5.52 5-77 0.25 Lower-middle 5-43 5-31 —0.12 Skilled working 5-33 5.20 —0.13 Unskilled working 5.20 5.16 —0.04 Work status Paid work 5.42 5-54 0.12 Looking after home 5-53 Marital status Cohabiting 5.04 5-^9 0.25 Single 5-13 5.21 0.08 Separated 5.20 5-27 0.07 Divorced 5.26 5-17 —0.09 Widowed 5.26 5.61 0.35 Married 5-51 5-56 0.05 Children No children 5.16 5.21 0.05 At least one child 5-44 5-47 0.03 Religiosity Attend service every week 4-99 5.02 0.03 Never attend 5-91 6.08 0.17 Gender equality Traditionalist 5.90 5-97 0.07 Egalitarian 5-49 5-51 0.03 Religion Catholic 5-68 5.67 —O.OI Protestant 5.63 5-79 0.16 Orthodox 5.38 5-4 0.02 Jewish 5-74 5-51 —0.23 Muslim 5.67 6.08 0.41 All 5.41 5-44 Note: The mean position of women and men on the ten- point Left-Right Ideology Scale, coded from left (low) to right (high) rpQpnť*; wnmpn mnrp rnnservafive tV In the "difference an mpm a nnsirive ' column, a negative niTlirp rpnrpcpnf« wi figure rep-impn mnrp left-leaning than men. Source: Pooled WVS/EVS, 1995-2001. 93 94 II- The Consequences of the Rising Tide clustered in the center of the ten-point scale. Nevertheless, some gender differences are evident; women are slightly more conservative than men among the less educated, the oldest (60+) generation, those looking after the home rather than in the workforce, the lower-middle or working classes, the divorced, and those of Catholic or Jewish faith. By contrast, men are slightly more conservative than women across all other categories, with this pattern clearest among Muslims, those with a university education, and the under thirty age group. If we extrapolate from these preliminary findings to the broader social trends that we have already observed, it suggests that more left-leaning women are found among the younger generation of well-educated working women, a group that has expanded in recent decades. Not all of the results point in a consistent direction, but if these patterns persist, it suggests that, owing to trends in educational and employment opportunities and the long-term process of generational turnover, women will probably gradually drift further to the left of men in the future. Nevertheless, these preliminary findings could be spurious, and we have not yet explored the impact of cultural values. For a more comprehensive multivariate analysis, we will use ordinary least-squares regression models with the ten-point left-right ideological scale as the dependent variable, again using the pooled 1995-2000 WVS/EVS across all nations. Pooling the data assumes that ideological positions should be relatively stable, so that there will be minimal change over the last two waves of the surveys. There may be some short-term fluctuations in particular countries - for example, those caused by a change of party government - but these effects would tend to cancel out across many societies. A negative coefficient in the models indicates that women are more conservative than men; a positive coefficient indicates that women place themselves to the left of men. In Model 1, we explore the impact of gender on the left-right ideology scale without any controls. This indicates the direction and significance of the direct effect of gender on political ideology. Model 2 then adds structural controls, including age, education, religiosity, class, labor force participation, marital status, and trade union membership, in that order, for the reasons already discussed. (See Appendix B for full details of the items and coding.) Model 3 adds cultural factors, measured by the Gender Equality Scale already used in Chapter 3, support for The Gender Gap in Voting and Public Opinion 95 postmaterialist values, and a scaled battery of two items measuring attitudes toward the role of government. Our analysis is designed to determine whether gender remains a significant predictor of political ideology after controlling for these factors. The results of the full models are given in detail in Table 4.4 for the pooled sample in all societies; Table 4.5 compares the results of the full model for postindustrial, industrialized, and preindustrial societies. Model 1 in Table 4.4 indicates that gender does predict ideology, with women leaning left, confirming the existence of the modern gender gap. Differences between men and women are modest in size but statistically significant. Model 2 shows that gender remains important even after controlling for social structure - in fact, the gender coefficient becomes stronger. As already observed, the effects of gender interact with age. The results of the model suggest that the modern gender gap cannot be explained, as some previous research suggests,46 as simply the result of differences between women and men in religiosity, class, age, marital status, or participation in the labor force. In postindustrial societies, the pattern of the ideological gender gap by age group is an important indicator of generational change: among the youngest group, women are more left-leaning than men, while among those over sixty-five the gender gap reversed, with women more conservative. In view of the long-term process of population replacement, this pattern may have important future consequences, moving women gradually further left. The other social variables behave in the expected manner, with stronger conservative values being evident among the older generation, the religious, and the middle class; by contrast, trade union membership, employment, and education were associated with more left-leaning positions. Lastly, Model 3 adds the attitudinal variables, which reduce the effect of gender, which nevertheless remains significant. In the final model, support for postmaterialist values, the Gender Equality Scale, and the Role of Government Scale are among the strongest predictors of left-right ideology, along with religiosity, although the structural variables also remain significant, with the exception of age. This final model suggests that the modern gender gap is more strongly influenced by culture than by social structure.47 In other words, women have moved to the left ideologically primarily because of a broad process of value change; the shift towards more egalitarian attitudes associated with postmaterialism and sex roles are o o N I Os On Üs ■Si» «a S O S 6« V) ŕ 4 4 H ►J M < rr\ a "« U T3 O + O o u « s + U c o .. U " o "S Sh 'S-a u ca Ž! ON O N O o M (M i-v. 0\ i^ H M O O H O M Q I I Ž í ^ w r»N H <^ H ft K ^ "~> o H -> O 1 H O o 1 o 1 O 1 H PA "t- t\ I I I £0^ W5 4J •-H TO U <_) O vi « *o -^ ° ft, o Oj ' "-» (U O Hl ^ 0 — « .2 -S u £S .2^ K, 3 — 32 U IUI IIÍI Hl HU liSI 111" Ulli Uši 1! ilíi i!« M lt.il rt íU a. _o <-i _o c<> T^e Gender Gap in Voting and Public Opinion 97 table 4.5. The impact of gender on the Left-Right Ideology Scale by type of society, Xgyf-2001 A grarian Industrial Postindustrial B Beta Sig. B Beta Sig. B Beta Sig. Male Gender (1) .221 .041 *** .071 .016 » .092 .024 ** Social structure Age (years) .002 .011 —.002 —.010 .002 .019 Education (three -.048 —.014 .058 .024 ** categories) Religiosity .015 .138 **» .013 .160 *** .009 .141 *** (100-point scale) Middle-class (i/o) .408 .071 *** .067 .014 * .282 .074 •** In paid -.665 —.119 *** -.080 —.017 * .004 .001 employment (i/o) Married or —.170 —.031 *** —.026 —.005 .092 .023 ** cohabiting (i/o) Union member (i/o) —.201 —.031 **« -.145 —.026 *** -.068 —.016 Cultural values Gender Equality —.021 -.138 «* -.008 -.066 *** —.013 —.109 *** Scale Postmaterialism —.072 —.032 *»* —.090 -.048 *** -.274 —.173 *** Role of Government .004 .034 *** .010 .092 *** .017 .164 *** Scale Constant 6.09 5.24 5.11 R .251 .215 .352 Adjusted R2 .062 .046 .123 Noře: See Table 4.4 for details. Source: Pooled WVS/EVS, 1995-2001. important components of this process, although structural changes also contribute to it. Does this result apply to both rich and poor nations? To address this question, Table 4.5 analyzes the final regression model across three types of society. Similar patterns for the ideological gender gap are found across all three models. Gender is significantly related to ideological position, with women more left-leaning than men in all types of society, even after controlling for gender differences in social structure and in political attitudes. The strongest model, explaining the most variance in the left-right ideological scale, is the one analyzing the postindustrial nations. There are many common structural 98 IL The Consequences of the Rising Tide factors, with religiosity and a middle-class occupational status being consistently associated with more conservative values. But the cultural variables prove to be the most significant and consistent predictors of positions on the ideological scale across all types of society; people who adhere to the values of more egalitarian sex roles and support for postmaterialism are more left-leaning, while those who prefer a minimal role for government are more conservative in their ideological position, in rich and poor nations alike. Conclusions and Discussion When women were first enfranchised it was anticipated by many contemporary observers that they would vote as a bloc, thereby transforming party politics. Although this did not happen, even modest gender differences in the electorate have often proved to be significant, and sometimes decisive, for political outcomes. The modern gender gap is now an established feature of the American political landscape, although, despite the extensive body of research on the topic, the reasons for this phenomenon are still not generally understood. Much of the previous literature has attempted to explain the American gender gap on the basis of factors specific to U.S. politics, such as party polarization over issues such as the ERA or the highly polarized politics of abortion and welfare reform.48 If, in fact, the root cause of this phenomenon reflects broader structural and cultural trends common to postindustrial societies, as we suggest, we would expect to find similar patterns of gender realignment emerging in other postindustrial societies. Commenting on the comparative evidence from the early 1970s, Inglehart argued that gender realignment in advanced industrialized societies was likely to occur in the future: We might conclude that sex differences in politics tend to diminish as a society reaches an advanced industrial phase. Or, going beyond our data, one could interpret the cross-national pattern as reflecting a continuous shift to the Left on the part of women: in the past they were more conservative than men: in Post-Industrial society, they may be more likely to vote for the Left. The relative conservatism of women is probably disappearing.49 Norris's research based on the 1983 Eurobarometer expressed cautious views about the existing pattern of gender differences in Europe at The Gender Gap in Voting and Public Opinion 99 the time, but also speculated that the conditions might prove ripe for change: We can conclude that there was no voting gap in European countries in recent years; overall women and men were very similar in their electoral choices and ideological positions. There is a potential gender gap, however, as women and men disagree significantly on a range of issues. These policy differences have not yet translated into voting differences, but they could, given certain circumstances.50 By the 1990s, these predictions seem to have been confirmed. The theory developed in this book argues that common developments transforming the lifestyles and values of women and men in postindustrial societies have altered political values and electoral preferences. This analysis points to four main findings. 1. In most nations today, women hold more left-leaning values than men in their attitudes toward the appropriate role of the state versus the market, favoring active government intervention in social protection and public ownership. z. Although an extensive body of evidence indicates that from the 1950s to the 1980s, women tended to be more conservative than men in their ideology and voting behavior, this pattern has now changed, with women becoming more left-leaning than men in many societies. 3. In explaining this phenomenon, we have demonstrated that the modern gender gap persists in many nations even after introducing a battery of social controls, but that the size of the gap diminishes substantially when we take cultural values into account. This suggests that the modern gender gap reflects differences in the value orientations of women and men, especially in their attitudes towards postmaterialism, the role of government, and gender equality, more than differences in their lifestyles and social backgrounds. 4. Lastly, analysis of generational differences points to intergenera-tional value change in postindustrial societies. The modern gender gap in ideology is strongest among the younger age groups, while the traditional gender gap persists among the elderly. If this finding reflects a generational change, as seems likely, rather than a life-cycle effect, it implies that in the long term, as younger ioo IL The Consequences of the Rising Tide voters gradually replace older generations, the shift toward left-leaning values among women should become stronger in affluent nations. These results have important implications for understanding the emergence of the modern gender gap in the United States. The evidence indicates that the realignment in the United States is not sui generis, but instead represents an enduring gender cleavage that is now becoming increasingly evident in other postindustrial societies as a result of long-term value changes. The gender gap in the United States has served to increase media attention and public debate about gender issues, to heighten party competition to gain the "women's vote," and to increase incentives for parties to nominate women for public office. The extent to which these developments will occur in other political systems remains to be seen. In order to gain electoral clout, women as a group need to participate at least as much as men, and to show a distinctive political profile. If they do, then these ideological differences promise to have significant consequences for the future power of women in representative democracies. The following chapter examines whether women are becoming increasingly active in public life, particularly through electoral turnout, civic engagement, and participation in new social movements. 5 Political Activism The earliest studies of political behavior in Western Europe and North America established gender as one of the standard variables routinely used to explain levels of electoral turnout, party membership, and protest activism, alongside the most powerful predictors of age and education.1 Based on a seven-nation comparative study of different dimensions of political participation conducted during the 1970s, ranging from voting turnout to party membership, contact activity, and community organizing, Verba, Nie, and Kim concluded: "In all societies for which we have data, sex is related to political activity; men are more active than women."2 During the same era, Barnes and colleagues (1979) found that women were also less engaged in unconventional forms of participation, such as strikes and demonstrations.3 The literature suggested that the well-established gender gap in many common forms of political participation remained evident during the 1980s and early 1990s in many countries around the world - even in the United States and Western Europe, where women have been enfranchised with full citizenship rights for decades.4 Nevertheless, given all the other substantial changes in women's and men's lives that have already been documented, we would expect to find evidence that some of these gender differences have gradually diminished or even disappeared over time, with women becoming more active, especially among the younger generations in affluent modern societies. To examine the broader picture, this chapter starts by comparing patterns of traditional political activism via elections and parties; civic 101