The End of Marriage? Individualism and Intimate Relations Jane Lewis Barnett Professor of Social Policy, University of Oxford, UK Edward Elgar Cheltenham, UK • Northampton, MA, USA introduction 15 THE EMBEDDED INDIVIDUAL At the micro-level of the family, marriage in the twentieth century certainly appeared tobe an individual choice. Yet as Regan (1999, p. viii) has observed, it also 'generates connection whose significance is not fully captured by the idea of consent'. In Regan's analysis, the neo-classical economist's view of marriage fails to take account of this sense of connection in the form of attachment to, identification with and commitment to care for others that is also highlighted in feminist writing on the family (Gilligan, 1982; Held, 1993; Griffiths, 1995), The notion of connection makes marriage a community with associated obligations. The literature on marriage has always exhibited considerable confusion as to how to characterise the nature of the marital relationship, whether in terms of two individuals who enter into a contract, or as a status, or as a sacrament. Late twentieth-century socio-legal literature in particular has been divided on how far to treat marriage as a contract pure and simple. Do intimate relationships consist merely of two independent individuals who agree to an association for whatever reason, or does the sum of the relationship amount to more than its parts, and if so, is this more true of marriage than cohabitation? Intimate relationships arc socially embedded. In the first place, intimate association results in more than just the satisfaction of the interests of the individual. After all, divorce would not be so painful if there were no sense of connection and attachment. Furthermore, expectations of intimate relationships in respect of both the individual and 'the relationship' arc embedded in wider social structures. Changes in these may affect the balance between the concern for self and other and are iikely to provide the context for understanding both decision making and tensions in family life. This book pays particular attention to two of the issues that have been most discussed as being part and parcel of family change: changes in the gendered divisions of labour, and in the law governing marriage and divorce, both of which historically helped to enforce traditional ideas about the role of men and women in marriage. In the main, the argument that we are seeing increasingly selfish individualism on the parL of adult family members has been deduced from the statistics on family change and associated with particular behaviours, in particular, the growth in women's employment, aided and abetted by legal change. There have certainly been major shifts in thinking and behaviour at the individual level and in law, but it is very difficult to understand the reasons for them and the nature of the causal relationships. Increased female employment, for example, affects not just women's own choices but the whole balance of activity in the family. The male breadwinner model family not only described a dominant pattern of behaviour, but also provided an accepted framework for marriage. As much critical feminist analysis has pointed out, the unwritten 'marriage contract' (enforced only at the point of breakdown) relied on the gender-specific roles that 16 The end of marriage? were assumed by the model. Changes in women's economic behaviour are therefore likely to carry implications for what has been a major prescriptive framework in respect of marriage and the family, but how and with what effects has not been explored. Similarly in the case of legal change, the strict external moral code that encompassed traditional gendered patterns of behaviour in the family and underpinned family law has been eroded. This pattern of change has been complicated in terms of cause and effect in relation not just to behaviour, but, again, normative prescription. Law has an expressive function (Sunstein, 1997) and cultural as well as behavioural consequences (Pildes, 1991). Again, these aspects of change need to be explored: legal reform may be a response to changes in normative expectations as well as behavioural change, and may seek to consolidate or modify social norms as well as to recognise changes in behaviour. The Erosion of the Male Breadwinner Model The traditional male breadwinner model family has described, with varying degrees of accuracy, gendered patterns of paid and unpaid work between men and women at the micro-level of the family during the twentieth century. It has also been as or more important in prescribing the normative expectations of men and women in families as to how their affairs should be arranged, probably for a majority of couples for the first three-quarters of the century. Its erosion is fundamental to understanding family change, but it is nevertheless very difficult to interpret. Whether academic or popular, the literature drawing attention to increased individualism focuses on individual actors, in terms of their behaviour and/or mentalities, and in particular to the importance of women's increasing labour market participation. However, the meanings of these are not always what they seem. For instance, Giddens (1992) has argued, albeit without empirical investigation, that-individualism renders personal relationships more democratic, which is as plausible as the idea of individualism as a corrosive influence. In their economic analysis of divorce, Becker et at (1977) took the fact that the majority of petitioners for divorce are female as evidence of women's growing economic independence and autonomy. However, there are many other explains ions for the numbers of female petitioners. Men tend to react to breakdown wilii violence or by walking out of the relationship, leaving women to seek divorce (Phillips, 1988). As the economically weaker partner, women usually need to try to get the financial arrangements settled (Smart, 1982; Maclean, 1991). In other words, the meaning of petitioning for divorce cannot be assumed. It may be that their increased participation in the labour market has only an indirect effect on women's marital behaviour. Cherlin (1981) argued that women's increased labour market participation had not caused an increase Introduction 17 in divorce, but rather had facilitated it. Similarly, de Singly (1996) suggested that it is not women's employment per se thai is responsible for their greater readiness to consider divorce, but rather the awareness it creates of tensions with the marriage. This indicates the importance of considering the complex interplay between mentalities and behaviour at the level of the family, but it is also important to locate decision making at the micro-level in the wider context of structural and normative change. The concept of 'individualisation' endeavours to locate the choices made by individuals. It refers to the way in which people's lives come to be less constrained by tradition and customs and more subject to individual choice, which in turn can only be understood against the background of changes in the labour market and in social provision by the modern welfare state in particular (Beck and Beck Gernshcim, 1995; Kohli, 1986; Buchmann, 1989). Elizabeth Beck Gernsheim (1999, p. 54) has described the effects of individualisation on the family in terms of 'a community of need' becoming 'an elective relationship'. Elias (1991, p. 204) expressed a similar idea in the following: The greater impermanence of we-relationships, which at earlier stager; o lie n had the lifelong inescapable character of an external constraint, puts all the more emphasis on tlie I, one's own person, as the only permanent factor, the only person with whom one must live one's whole life. The family used to be a community o^ need held together by the obligations of solidarity. But women's increased labour market participation has resulted in a new division between biography and family. Burns and Scott (1994) have made a similar point in their discussion of the way in which male and female roles in the family have become 'decomplerncntary'. In essence, it is difficult to mesh the labour market biographies of two adults with family life. Individualisation thus pulls men and women apart (although Beck and Beck Gernsheim, 1995, argue that they are also pushed together again in new relationships, because as traditions become diluted, so the attractions of a close relationship grow). This kind of analysis is broad-brush. But it does highlight the importance of the effects of the decline of the traditional male breadwinner model family -which has been underpinned by both increased female labour market participation and the safety net provided by the welfare state - on intimate relationships. However, it is hard to generalise about any part of the developments described by Beck and Beck Gernsheim, while the assumptions made by those who assume that female labour market participation has resulted in female economic independence are suspect. In the UK, variations on the theme of a 'one-and-a-half-breadwinner model' have emerged, whereby women work part-time (often for very few hours) and use a variety of child-care arrange- 18 The end of marriage? ments involving husbands and other kin at the informal level, as well as publicly and privately provided services. The issue is what these patterns mean in terms of what they signal about the priorities of the men and women involved and their impact on the stability of the family. Given the degree to which women continue to be economically dependent on men, it is unlikely that all partnered women with children who are in the workforce are motivated by self-development and self-interest. Indeed, Catherine Hakim (1996) has suggested that there are two populations of women workers: a minority who want a full-time career and a majority who continue to put home and family first. However, in the British context it is likely that structural constraints in the form of lack of child-care also play a prominent part in explaining female labour market behaviour (Joshi and Davies, 1992). Nevertheless, a majority of women may be responding primarily to the dictates of the family economy and supplementing it to whatever extent is deemed necessary. This is one of the oldest established determinants of women's paid employment (Tilly and Scott, 1975), and may point more to the economic /«^-dependence of family members than to autonomy and independence. . Simple calculations at the individual level of economic self-interest are unlikely to be sufficient explanations. As a result of their examination of the position of British lone mothers, whose labour market participation declined in the 1980s, Duncan and Edwards (1999) have suggested that women do not abide solely by the dictates of economic rationality in making decisions about marriage, divorce, motherhood and employment. If we take the question: are women who work full time and earn a reasonable wage more or less likely to marry and/or divorce? part of the answer will lie in the domain of economic behaviour outside the family, but part will also depend on gendered divisions inside the household, where the key factor may be the woman's priorities and preferences, or the man's selfishness, irresponsibility or mere inflexibility. Arlie I iociischild's (1990) interview data have documented the way in which women have experienced difficulty in dealing with the 'double shift' that has resulted liom llie fact that, while they have increased their labour market participation, II iťii have noL significantly increased their contribution to unpaid work. Ferri and Nmilh's (1996) survey data have shown the extent to which British men arc conient with such a pattern and women arc not. li may be, as Oppenheimcr (1994) argued, that with the decline of the male breadwinner model the whole meaning of marriage and partnership has changed. (iiven that two incomes are now necessary to meet household expenditure (particularly mortgages), career women may be more marriageable and may help to secure family stability rather than jeopardise it, pace Becker (Oppenheimcr and Lew, 1995). Oppenheimer (1994) has suggested that instead of positing the old specialisation of men as breadwinners and women as housewives and Introduction 19 carers, relationships are now based on the ability of each partner to make a contribution, unique or similar, and to pull his or her weight. If this constitutes a new set of assumptions about behaviour in intimate relationships, then again women's employment will not necessarily pose a threat, although male selfishness and/or inflexibility may well do so. Indeed, it is possible thai women's employment may affect fertility rates more than divorce rates. But in respect of understanding the causes of change in marriage and the family, the point is that it is not enough to consider individual choices (which may or may not be driven by selfish individualism) even when they are located in their structural context, as Beck and Beck Gernsheim strive to do. It is also necessary to pay attention to assumptions as to what is considered to be appropriate behaviour. Value systems (shared goals) and norms (shared goals specifying functions, that is, rules regarding the 'ought' supported by sanctions2) are a crucial part of the context of the individual actor. They are part of both cause and effect in respect of increasing individualism. The erosion of what I have called the normative prescriptions which accompanied the established, traditional economic and legal frameworks that dominated marriage and family life for the first three quarters of the twentieth century, in the form of the male breadwinner model and the imposition of an external moral code, has left a vacuum. Oppenheim Mason and Jensen (1995) have drawn attention to the way in which the micro-theory of the neo-classical economists does not recognise or allow for collectively generated or agreed upon norms and sanctions. Thus culture often remains absent from, or peripheral to, explanation. For example, while Fukuyama (1999) acknowledged the importance of culture, arguing that it allowed Japan and Korea to stave off 'the great disruption', he could see no means of explaining why 'attitudes' in western countries should have begun to change rapidly in the 1960s unless they were driven by something else. He preferred therefore to explain family change in terms of social and economic variables: namely, the increased use of artificial birth control and women's greater labour market participation. These in turn changed the norms that constrained men's behaviour and allowed them to behave more irresponsibly. Thus, Fukuyama's explanation prioritised social and economic behaviour; attitudes and norms played a secondary part in the story. He may well be correct, but his relegation of cultural factors to second place meant that, as is commonly the case, little attention was paid to the precise way in which they worked. It is noteworthy that those who successfully demolished the argument that state welfare benefits had 'caused' ione motherhood concluded at the end of empirical work that considered only quantifiable variables and found them wanting, that 'our hunch is that the real force behind family change has been a profound change in people's altitudes about marriage and children' (Bane and Jargowsky, 1988, p. 246). Inglehart (1997) has argued thai while major divorce 20 The end of marriage? law reform happened suddenly in most western countries {at the end of the 196ÜS and during the 1970s), it followed a long process of attitudinal change. However, Oppenheimer (1994) has pointed out that American panel data show that women's changing attitudes towards divorce between 1962 and 1977 had no significant effect on whether their marriages broke down. Rather, the strongest determinant of attitudinal change was whether a woman experienced divorce or separation in that period. It is no more easy to trace a causal relationship between changes in attitudes and individual marital behaviour than between the latter and women's increasing capacity for a measure of economic independence. Sociologists are divided, for example, on whether increased female employment is linked to more egalitarian gender role attitudes (Haller and Hcllinger, 1994; Alwin, Braun and Scott, 1992). Furthermore, the relationship between attitudes and values is far from simple, just as in the case of the relationship between behaviour and attitudinal change. Van Deth and Scarborough (1995) argue that attitudes are empirically measurable whereas values are not. indeed, value systems may stay the same while behavioural norms change. Thus people may continue to believe in the importance of traditional family life, and possibly express attitudes in keeping with this, and yet divorce (Scanzoni et ai, 1989). Norms in respect of family behaviour have undoubtedly changed; the stigma attaching to divorce and unmarried motherhood has eroded in the post-war period, slowly at first and rapidly from the late 1960s. Rules regarding the order in which people have sex, have children and marry are no longer firm and there are few informal .sanctions (formal sanctions in the form of law do exist; for example, cohabiting fathers do not have the same rights in respect of their children as married fathers, but people are often blithely unaware of such distinctions). Nevertheless the vast majority of people still say that they value family life (Scott, 1997). A major divide in the literature is between those who see culture as integral to explanation and those who view it as exogenous. Rational choice theorists such as Jon Elster (1991) may acknowledge the existence of norms, but use them only as a residual form of explanation, to be invoked to explain the awkward bits that are left when the rational choice analysis is completed. Sophisticated economic approaches, such as that using the idea of the 'convention', interpreted by Sugden (1998a, p. 454) in terms of tacit agreements or common understandings giving rise to shared expectations, acknowledge the importance of the cultural variable, but stop short at the idea that norms are internalised. Sugden has argued that conventional practices can generate normative expectations, which may in turn be significant for the stability of conventions. Others insist that norms and values are embedded in society and are part of the framework within which choices are made. Norms by definition are not chosen, and a decision to abide by them may be made consciously or seemingly without any conscious interrogation of alternatives.3 Thus Sunstein (1997) has insisted Introduction 21 that individual choices arc a function of norms, meanings and roles, and thai individuals may therefore have little control over them. Puzzles of rationality, he argues, arc the product of social norms and moral judgements. Processes such as individualisation are linked to norm-change. The male breadwinner model did not merely describe a pattern of economic activity in the family. Because it was internalised by a majority, certainly in the immediate post-war years, it also conditioned expectations within marriage. The decline of the model in respect of the changing labour market behaviour of women might be expected at some point to have been accompanied by a shift in normative expectations on the part of people and legislators, which may turn out to be more radical than the actual change in behaviour. As Lessig (1996, p. 285) has observed, when norm violation increases, then the meaning of obeying the norm also changes and at some point 'obeying the norm makes one a "chump1". Stacey (1990) commented that young working-class men in late twentieth-century America were not sure whether to regard one of their number who becomes a breadwinner as a hero or a chump. This is in large measure because normative meanings and expectations are far from clear. The norm may now be that women will engage in paid work. Attitudinal surveys have shown consistent increases in the acceptance by men and women of female employment. But to what extent - full-time or some form of part-time - is not entirely clear. Nor is it clear what the accompanying assumptions are in respect of unpaid work. Beck and Beck Gernsheim (1995) have suggested that it is no longer possible to know what marriage and family life mean. The norm is no longer unambiguous, and this opens up a space for negotiation. This is where the importance of the link between changing structures and norms at the collective level and changing mentalities and behaviour at the individual level begins to become apparent. In regard to the increase in women's employment, which has been identified by so many as important in explaining family change, the mechanisms by which this happens are hard to elucidate. In all probability there are many, but one promising line of inquiry would seem to involve the way in which the increase in women's labour market participation helps to effect a shift in the male breadwinner model, which in turn shakes the whole fabric of gender roles that have been widely assumed - by government and by people - to flow from it. Lesthaeghe (1995) referred to the importance of the erosion of normative prescription when he stressed the significance of the emergence of a sense of individual autonomy, which in his view had nothing to do with egocentric behaviour, but meant only that the individual no longer takes externally supplied norms and morality for granted, relying instead on his or her own judgement. At some point, the gap between practices and the normative expectations flowing from the male breadwinncr model becomes too great. The rupture gives rise to a new set of normative expectations. Thus, female employment is expected, although to what degree 22 The end of marriage? is unclear. Nevertheless, expectations may actually ran ahead of behavioural change, aided and abetted by changes in social policies. Thus at the end of 1990s, the British government swung towards treating all women, including lone mothers, as paid workers for the purposes of the social security system, regardless of the fact that a majority of mothers work only part-time. It is possible to suggest that late twentieth-century marriages, based on romantic love and increasingly bereft of the prescriptions that accompanied major normative frameworks, such as the male breadwinner model family, became not just more fluid in the sense of movement in and out of them, but also more open to negotiation in terms of their organisation. As part of a theoretical model for comparative gender research, Pfau-Effinger (1998) has suggested that negotiation takes place when a gap opens up between the gender culture (in the form of norms) and the gender order (that is inscribed, for example, in social policies and labour markets). It may be, as so many insist, that behaviour has driven social change in respect of the erosion of the male breadwinner model, but it is in all probability the change in 'normative expectations' that has opened up the space for negotiation. Once normative expectations change, they will cease to reinforce the behavioural dimensions of the male breadwinner model. But negotiating the gap that results is not easy and at the level of the couple breakdown may occur. At the level of policy making, new problems arise when governments begin to make assumptions based on a more egalitarian set of normative expectations about women's paid work that may be almost as out of touch with the social reality as the old . assumptions that adult women were housewife/carers in the male breadwinner model family. The Erosion of an Externally Imposed Moral Code Defenders of a strict law of marriage and divorce have usually regarded it as a means of stopping people from exercising their individual preferences regarding the formation of intimate relationships and thus threatening family stability. But how far family law could be used to uphold the discipline and order of marriage became an increasingly contested issue, first in the 1920s and again in the 1960s. The relaxation of the divorce law followed (in 1937 and 1969), hul even at the end of the century, neither cohabiting nor married people were I'u H y individualised in the eyes of the law. Changes in family law, which has historically been based on an externally imposed moral code, have been identified alongside changing labour market behaviour as being particularly significant for understanding family change. Irene Thery (1994), a French sociologist, has captured the nature of the relationship between change at the individual level and legal change in a rather different way, using the idea of 'demarriage', by which she means the disen- íntroduction 23 gagement of the institution of marriage from other social structures, such as the law. This idea has some parallels with Beck and Beck Gernsheim's concept of individualisation and gives meaning to the oft-cited idea among sociologists of marriage moving from institution to relationship (proposed first by Burgess and Locke, 1953). Marriage obviously remains an institution, but it is important to realise the extent to which it has broken free of major prescriptive frameworks and in that sense has become more of a private relationship.4 The process of 'demarriage' does not necessarily affect the number of people who choose to marry, but makes it easier for people to move in and out of marriage. Whereas the vast majority of people's lives only a generation ago looked rather simple in terms of family patterns - marriage followed court.ship and death ended marriage - today the patterns are increasingly complicated. The vast majority cohabit before marriage and one in three divorce. These may go on Lo cohabit again and/or marry and possibly divorce again. Marriage has become a discretionary adult role (Bumpass, 1990), whereas it used to be compulsory if certain other things were to be achieved: a home of one's own and children. Again, it is questionable whether the effect of legal reform on family change, that is, on behaviour, has been direct. Dror (1959) concluded that legal changes had more direct impact on emotionally neutral and instrumental areas of human activity than on expressive and evaluative arenas such as the family. In respect of divorce in particular, Rheinstein's (1972) comparative socio-Iegal study argued that marital breakdown could be high, even in the absence of divorce. Similarly in his comparative historical study of divorce, Phillips (1988) insisted that levels of marital breakdown have not been dependent on legal change. However, Rheinstein called attention to the importance of what he called the 'cultural climate', arguing that the incidence of breakdown increased in times of accelerating social change 'with the unsettling tendencies towards anomie' (p. 311), thus making the connection between marital behaviour, the socioeconomic context and mentalities. It is also likely in the case of family law that legal change not only contributed to a shift in behaviour but was also related to a shift in normative expectations. Weitzman (1985) suggested that when the rules were changed regarding what was expected of husbands and wives on divorce, so this also changed norms in respect of marriage. Like social policies in respect of cash benefit entitlements, it is likely that family law facilitates rather than causes behavioural change at the individual level, but, because it has historically been underpinned by an external regulatory moral framework, relaxation of divorce law may serve to legitimate changes in behaviour. Phillips (1988, p. 617) concluded that 'it seems probable that divorce does breed divorce', because it becomes part of the cultural climate within which marriages exist and hence the stigma once associated with it disappears. Hirshleifer (1998, p. 301) has pointed out that a social convention, such as the idea that cohabitation should occur only after marriage, may change partly 24 The end of marriage ? as a result of conspicuous violations by individuals. At a different level again, if the parents of a cohabiting couple fail to impose sanctions, then other parents may conclude that this is the way that they too will respond. However, it is possible that prior impetus to breaking the expected and accepted pattern of behaviour in regard to cohabitation may have come from public debate and opinion-formers. At some point, possibly before, but more likely after major behavioural change of this kind, official sanction may be given in whole or in part by legal reform. But the relationship between law and social norms is also complicated. The state may impose law from the top down or endeavour to enforce social norms from the bottom up (Cooter, 1997); it may serve to strengthen or weaken norms (Posner, 1996); or it may transform them (Lessig, 1998). The nature of the law is part of the socialisation process and one of the means by which norms are internalised (Dau Schmidt, 1997). The state plays a major role in securing the stability of norms (by imposing legal sanctions) and may also either respond to a behavioural change that instigates norm change or itself act as an instigator by enacting legal reform that actually runs somewhat ahead of behavioural change. For example, divorce law reform in Britain in 1984 adopted the 'clean break' approach, which treated men and women as if they were fully individualised and economically independent when for the most part they were not. As in the case of the erosion of the male breadwinner model, it seems that more attention should be directed towards the importance of cultural changes. These worked'out rather differently in the case of family law, where radical changes in thinking on the part of elite opinion about the desirability and possibility of imposing a. strict external moral code played a major part in securing the reform of the strict moral code embodied in the law. But it may be argued again that the erosion of this form of prescription opened up a space for negotiation at the individual level, particularly in respect of whether to marry. Legal reform that resulted in the progressive relaxation of the divorce laws in particular has been widely characterised as part of a process of 'deregulating' family iaw. Glcndon's (1981) thesis on deregulation has been most influential. She argued that while 'the legal ties among family members are becoming loosened, the web of relationships that bind an individual to his job (and his job to him) is becoming tighter and more highly structured' (p. 1). This analysis of deregulation in the private sphere and more regulation in the public world of work and welfare did not look so convincing by the 1990s, by which lime the restructuring of welfare states had moved significantly towards emphasising personal responsibility, while labour markets had become more 'flexible', with more short-term contracts and part-time hours. Indeed, reform in the field of family law was accompanied during the last two decades of the twentieth century by increased uncertainty in the public world of work and welfare. In fact, the idea of deregulation did not quite serve to capture the nature Introduction 25 of legal change. The process of liberalisation that began in Britain with the major reform of the divorce law in 1969 has increasingly allowed men and women as husbands and wives to order their own affairs, rather than imposing a strict moral code from above. Thus, family law has effectively assumed the existence of a degree of individualisation. However, it has also put in place measures (such as the 1991 Child Support Act) to regulate men and women as fathers and mothers. Thus during the last part of the twentieth cenlury the prescriptions of an external moral code embodied in family law and designed to operate in concert with the assumptions underpinning the male breadwinner model have been eroded at the same time as economic conditions have undergone major change, again affecting normative expectations regarding employment and the life course for women and men. Family matters have increasingly been left to individual adults, but, at the same time, family law has tried to enforce individual responsibility, which often continues to be defined in traditionally gendered ways. Habermas's (1996) volume on the theory of law and democracy captures what happens when an established moral code ceases to exist and morality is assumed to come from within: 'under the moral view point of equal respect for each person and equal consideration for the interests of all, the henceforth sharply focused normative claims of legitimately regulated interpersonal relations are sucked into a whirlpool ofproblematisation' (p. 97). Societal plu-ralisation and fragmentation mean that it is difficult to agree on an external moral code. But this does not mean that the vacuum was necessarily filled by simple deregulation. Rather, it may be argued that the increasingly popular view that morality should come from within was accompanied by a shift in the nature of regulation: from judging the affairs of the couple according to pre-set criteria, to making sure that the couple lakes responsibility for sorting out their own affairs. Many of those convinced as to the importance of the growth of individualism have tended in the main to read it off from the dramatic changes in the statistics recording family change, and then to link it either to changes in individual behaviour, such as women's increased attachment to the labour-market, or to changes in mentalities, such as prioritising personal growth. These studies do not in the main investigate what is actually going on inside personal relationships, even though they are preoccupied with change at the individual level. In addition, change at the individual level will necessarily be affected by what is going on outside the family, for example in the labour market or family law. However, the relationship may not be direct, but may rather be mediated by changes in normative expectations about behaviour, particularly in respect of the nature of the division of labour and the sexual moral code. It is not the project of this book systematically to review, eliminate and order all the possible variables feeding the idea of 'increasing individualism' as a 26 The end of marriage? cause of marriage and family change. Rather, it will seek to examine change in some of the key factors, focusing on the erosion of normative prescriptions emanating from the male breadwinner model and from family law, and on the changes that have taken place inside the relationships of married and cohabiting people, particularly in respect of the division of labour, using both contemporary studies and interview data. Section II of the book pursues the development of the idea of the male breadwinner model and how this, together with the external moral code which was embodied in family law and which served to underpin the gendered nature of obligations associated with the traditional family, have been eroded. It focuses in particular on the mediating effects of changes in prescriptive frameworks and their relationship to law and behaviour. The charting of the debates about marriage and the family serves the further purpose of putting the current anxieties about the growth of individualism into historical perspective. Section III reports the results of a qualitative study of the relationships of married and cohabiting couples with children, and of their parents, which investigated the meaning of obligation and commitment and asked whether cohabitants were more individualistic than married people and whether the younger couples were more individualistic than their parents. The research reported in this section demonstrates how the erosion of prescriptive frameworks has been experienced. As we have seen, the meaning of behaviour labelled individualist may actually be very different when it is contextualised and certainly calls into question the idea that individualism is necessarily selfish. Commitment at turn of the twenty-first century may lake different forms, but this does not necessarily pose problems for public policy. The final chapter of the book addresses the vexed issue of what might be done. Galston (1991) has argued that the liberal state must promote what it believes in: its own concept of the human good, which includes attention to moral issues. When traditional forms of prescription are eroded, it is not surprising that governments begin to talk more openly of morality, but they face the problem that in a pluralist society no one view of what constitutes moral standards will hold. Nevertheless, there is still a strong idea that the state should play a role in setting standards of behaviour. Mead (1997) has identified a more general trend in social politics towards a 'new paternalism', which, for example, proposes to enforce responsibilities and to force people drawing state benefits to take up opportunities, especially in relation to employment. However, il is also possible to argue for more non-intrusive state support for the family, or even for families. Both approaches could be found in British family law and policy in the late 1990s. Some have argued that the market mechanism of contract might be extended to what is perceived as the new individualistic world of the family, for example, through pre-marital and cohabitation contracts (Posner, 1992). However, this has met with considerable suspicion, chiefly on the ground that it would further erode trust (Anderson, 1993), which is an Introduction 27 argument that stems in part from a conviction that the private world of the family risks contamination by the public sphere. However, there are many other reasons to question whether the morality of contract is appropriate, not least in regard to the inequalities between the parties to it (Baier, 1986). Many of Ihe communitarians, such as Bellah, who were among the first to raise the alarm about selfish individualism, have expressed suspicion of both an overly attentive state and an unregulated market (Wolfe, 1989; Elshtain, 1995). Thus, Fukuyama (1999) has argued that only a spontaneous regeneration of norms regarding the importance of the traditional two-parent family will do, There is little agreement even on how to begin to address the issue of marriage and family change, or the roles, possibilities and limitations of family law and family policy. Behind the charge of selfish individualism is the fear that obligations and commitment of adult men and women to each other, and of both to children, have been eroded. There are (empirical) straws in the wind to suggest that this may be too pessimistic a judgement. Cancian's (Í987) study of love in America stressed the extent to which individualism was not necessarily antithetical to feelings of responsibility for the welfare of others. Many of the couples in her sample stressed the value of interdependence in the context of greater individ-ualisation. Pahl (1996) has emphasised the importance of distinguishing between individualism and individuality, which carries no taint of selfishness. Interdependent people can value personal growth, individuality, equality and morality that comes from within rather than one that is imposed from without, and yet still feel commitment to one another. Such commitment is not socially prescribed and is in that sense individualistic, although not, as Lesthaeghe (1995) has remarked, necessarily egocentric. The process of individualisalion and the erosion of prescriptive frameworks inevitably leave something of a vacuum in the sense that there is no longer a widely accepted 'ought7. Il is in this context that the relational aspects of marriage have come more to the fore, as the working out of intimate relationships has become a more private matter in the first instance. Thěry (1998) has suggested that the meaning of 'the couple1, whether married or cohabiting, has changed from '2 makes 1' -the classical idea of marital unity - to ' I + i makes 2\ Changes in women's status have certainly been a prerequisite for this change; if not equals, both parties to a relationship are now people to be reckoned with. Both adults expect a 'say'. Indeed, negotiation of the moral and economic aspects of marriage has increasingly been pushed down to the level of the individual couple rather than being prescribed as part of the 'institution1 of marriage. But this does not necessarily mean that people will behave selfishly. Academics from a variety of disciplines and political positions have found evidence that people do employ 'moral sense' (Wilson, 1993) and take the issue of their obligations to others seriously (Finch and Mason, 1993; Smart and Neale, 1999). Bauman (1993, 1995) has argued that deregulation and the removal of rules has not meant that 28 The end of marriage? people have abandoned responsibility for making ethical choices, just as Karst (1980) found that intimate relationships generate moral obligations even when law does not enforce them. Thus, just because divorce has become easier, it does not mean that moral debate and personal responsibility have been eliminated. Indeed, one of the most suggestive trends in recent empirical work has been the idea that while at the demographic level families and family building are becoming ever more diverse, there is convergence in terms of the negotiated nature of commitment and responsibility (Weeks et ai, 1999). This is, on the whole an optimistic picture of change, which the research reported in Section III of this book tends to support. The final section of the book argues that this finding has important implications for family law and family policy, and a particularly important message for those who seek to use the law to put the clock back. NOTES 1. 1 am grateful lo Kathleen Kiernan (personal communication). 2. The definition of norms is recognised to be elusive and varies between the disciplines. P'iske (1992) differentiates further between norms, which prescribe functions and constitute external constraints, and moral standards, which insist on a particular goal as an 'ought'. Cooler (1997), Ľllickson, R.C. (1998) and Posner (1996) provide useful reviews from the socio-legal perspective. Of particular importance for this study is the controversy over how far norms are internalised. I believe that they are and that is what explains the prescriptive power of something like the male breadwinner model (see pp. 45-71). 3. The process by which this occurs is complicated; see, for example, Suchman (1997). 4. Laslett (1973) defined privacy in terms of structural mechanisms that prohibit or permit observ-abilily in the enactment of family roles. My analysis of the erosion of prescriptive frameworks may be seen as a development of this early work, but Laslell's accompanying idea that as family privacy increased, so 'social control' in the form of legal regulation would decline has not proved lo be the case (see p. 24 in this chapter and Chapter 5). 2. Changing patterns: the decline of marriage and the rise of cohabitation Marriage has been part of the typical experience for adults throughout the twentieth century. During the first half of the century the proportion of women who were or who had been married never fell below 60 per cent, and reached 75 per cent in 1951 and 79.4 per cent in 1971. However, since then the percentage has fallen back again to 1950s levels. The story looks more striking slili when told in terms of first marriage rates (Table 2.1). Since the mid-1980s these have fallen precipilously and are now considerably lower than at the beginning of the century. The number of first marriages in the mid-1990s was the lowest recorded this century, despite a much larger population (Haskey, 1995), and the mean age at first marriage for men and for women was at its highest. During the early part of the century, marriage conferred a higher status on women than spinsterhood. In common parlance, a woman 'failed' to get married. This was particularly the case before the First World War, but marriage remained the normative expectation of women in all social classes. At the end of the century, marriage is still an important event, indeed the amount spent on the average wedding has increased substantially to an average of just over £10 500 in 2000. However, getting married is no longer a mailer of urgency. For middle-class women, especially in the period before the FirsL World War, marriage was virtually their only means of financial support because few occupations were open to them. Spinsters faced an often lonely and marginal life in their parents' home or in the households of male relatives. Cicely Hamilton, a leading feminist, wrote bitterly in 1909 of marriage as 'a trade' for women. Working-class women were employed in large numbers, but few commanded wages much above subsistence levels. Marriage has become more of an option in terms of material survival and status, while living alone or cohabiting no longer stigmatises those involved. Divorce and births outside marriage were relatively rare until the last 30 years of the century. Prior to 1914, divorce was largely confined to the middle and upper classes. Changes in the aid given to poor petitioners in 1914, together with the disruptive effects of the First World War, produced an increase in the divorce rate after 1918, but it was not until 1946 that legal aid became freely available and therefore not until 3951 that divorce petitions began to come from 29 30 The end of marriage? Table 2.1 First marriages by age and sex, England and Wales, 1901-97 Year All ages Mean age Thousands Rate3 Men 1901 233.9 59.6b 26.7 1911 251.7 61.8b 27.4 1921 290.6 63.5b 27.5 1931 285.5 59.7' 27.3 1941 357.1 78.3 26.9 1951 313.5 76.2 26.7 1961 308.8 74.9 25.6 197Í ■ 343.6 82.3 24.6 1981 259.1 51.7 25.4 1985 253.3 46.6 26.0 1991 222.8 37.0 27.5 1995 198.5 31.8 28.9 1997 188.3 28.4 29.5 Women 1901 240.6 57.4b 25.3 1911 257.5 , 58.7b 25.8 1921 290.4 55.2b 25.5 1931 294.1 54.7 25.4 1941 365.0 74.5 24.6 1951 319.2 76.3 24.4 1961 312.2 83.0 23.1 1971 347.4 97.9 22.6 1981 263.4 64.0 23.1 1985 258.1 58.2 23.8 1991 224.8 46.0 25.5 1995 198.5 40.1 26.9 1997 188.5 35.6 27.5 Notes: ;L i'er 1000 single persons aged 16 and over. ''Üuinqucnniii rales 1901-05, 1911-15, 1921-25 Sources: OPCS, Marriage and Divorce Statistics: Historical Series 1837-1983, FM2 No. 16 (1 xmdon: HMSO, Í995) Tables 3.2a and b, 3.5a and b, 3.7a and 3.7b; ONS, Population, Trends88 (Summer 1997) (London; Office of National Statistics, © Crown Copyright 1995) Table 23; ONS, Marriage, Divorce and Adoption Statistics 1987-1997, FM2; no. 25 {London; Office of National Statistics, © Crown Copyright 1999) Tables 3.7, 3.8 and 3.15. The decline of marriage and the rise of cohabitation 31 a cross-section of the population (Rowntrcc and Carrier, 1958). Prior to Ihr Second World War, working-class couples made use of judicial separation machinery rather than divorce, but the number of informal separations w;is undoubtedly much larger than the number that came to court (MacGregor, O.R, et al., 1970; Stone, 1990; Phillips, 1988). Table 2.2 shows that the divorce rate began to climb significantly in the late 1960s and trebled during the 1970s. Il climbed slightly again in the early 1980s, and has remained high and stable since, falling back slightly in 1997. The extra-marital birth rate increased dramatically during the Second World War, but was due mainly to marriages mat were planned not taking place due to conscription, wartime disruption and death. Table 2.3 shows that while there was a significant rise again in the 1960s, the dramatic increase has taken place since 1985. The high rate of teenage births in the UK, most of which take place outside marriage, plays a large role in keeping the total fertility rate well above the European average (Coleman and Chandola, 1999). Table 2.2 Divorce rate per 1000 married population, England and Wales, 1950-97 1950 2.8 1960 2.0 1965 3.1 1970 4.7 1975 ' 9.6 1980 12.0 1985 13.4 1990 13.0 1995 13.6 1996 13.8 1997 13.0 Sources: OPCS, Marriage and Divorce Statistics 1837-1983 Historical Series, FM2 no. 16, (London: HMSO 1995) Table 5.2; Marriage and Divorce Statistics 1S37-19S3, FM2 no. 21 (London: HMSO 1995) Table 2.1; ONS Marriage. Divorce and Adoption Statistics 1997, FM2 no. 23 (London; Office of National Statistics, © Crown Copyright 1999) Table 2.2. Many recent commentators have located the beginnings of the striking change * in family structure in the 'permissive' 1960s. However, the story is more complicated than this. There is evidence that sexual activity among the young increased during the 1960s and that the growing tise of the contraceptive pili from ihe beginning of the 1970s strengthened this trend (Black and Sykes, 19/1; Bone, 1986; Moore and Burt, 1982; Schofield, 1968; Farrcll, 1978). Thus sex was increasingly separated from marriage and increased sexual activity resulted 32 The end of marriage? in a rise in both the extra-marital and marital birth rates {Table 2.3). This contrasts both with the war years, when the extra-marital birth rate rose faster í han Lne marital rate, and with the last decade when the extra-marital rate has risen and (he marital rate has fallen. In the 1960s there was still a tendency for pregnant women Lo marry. A majority of births to women younger than 20 years old were conceived outside marriage in the 1960s, but the majority of pre-niariiully conceived births took place inside marriage. In 1969,55 percent of cxlra-niarilal conceptions were legitimised by marriage, 32 per cent resulted in Liliegiliinale' births and 14 per cent were aborted (the Abortion Act was passed in 1967) (Lewis and Kiernan, 1996). Table 23 Marital and extra-marital births per WOO women aged 15^44, England and Wales, 1940-98 Marital birth rate per 1000 Extra-marital birth rate per married women 1000 single, divorced and widowed women 1950 108.6 10.2 í 955 103.7 10.3 I960 í 20.8 14.7 1965 126.9 21.2 1970 113.5 21.5 1975 85.5 17.4 1980 92.2 19.6 1985 87.8 26.7 1990 86.7 38.9 1995 83.3 39.4 1998 82.3 40.3 Sources: Office of Population Censuses and Surveys (OPCS), Birth Statistics: Historical Series 1837-1983, Scries PM1 no. 13 (London: HMSO), 1987) Table 3.2b and c; OPCS, Birth Statistics: Historical Scries 1837-1983, Series FM1 no. 22 (London: HMSO, 1995) Table 3.1; ONS, Birth Statistics 1998, Series FM1 no. 27 (London: Office of National Statistics, © Crown Copyright I<)<)<)) Table 3, ll>, ll is perhaps therefore not so surprising that there was seemingly little panic about what amounted to a significant increase in the separation of sex from marriage. The Pact that a majority of pre-marital conceptions were legitimised ;ind thai divorce rales were low (Table 2.2) accounted in large part for a series n| optimistic statements about family stability (see, for example, Fletcher, 1966; (ioiv.r, 197 I). As late as 1976 Norton and Glick puzzled over the continuing rise in ilie divorce rate in the United States and speculated (hopefully) that as time went on people's expectations of marriage would become more consistent with experience. The decline of marriage and the rise nfcohahitation 33 Since the beginning of the 1970s, there have been marked changes in marriage patterns, such as older marriage and substantial declines in marriage rales (Table 2.1)- trends that continue lo the present; a dramatic rise in divorce rates that plateaued from the 1980s; and the emergence of widespread cohabitation (Table 2.4). From the late 1970s, the proportion of births outside marriage began to increase slowly at first and then rapidly throughout the 1980s, with signs of stabilisation in the early 1990s at about one in three of all births (Table 2.3). Kingsley Davis (1985) offered a series of tests as a means of demonstrating whether marriage was declining: the postponement of marriage; fewer people marrying; a smaller proportion of adult life spent inside marriage; and a rising preference for competing types of intimate relationship. All these tests now seem to indicate that there is indeed a decline in marriage. In fact, declining marriage rates and increased childbearing outside marriage have been inextricably linked to the growth of cohabitation. Kingsley Davis took the view thai cohabitation was not substituting for marriage. The British Social Altitudes Survey for 1994 showed lhat attitudes towards childless cohabitation had continued to relax; 64 per cent felt that it was all right for a couple to live together without intending to marry. However, a majority (57 per cent) still fell that people who wanted children should marry (Newman and Smith. 1997). Nevertheless, childbearing has increasingly been taking place in cohabiting relationships. As Meuleman (1994) has commented, it is now clear that cohabitation is both sequel and alternative to marriage. Cohabitation was apparently common in the early part of the twentieth century, when divorce was rare (Gillis, 1986). When separation allowances were provided for the wives of servicemen during the First World War, special provision had to be made for 'unmarried wives' (Parker, 1990). However, cohabitation was probably at its nadir in the 1950s and 1960s, when marriage was almost universal (Kiernan and Estaugh, 1993). Living together as a prelude to marriage began in the 1970s. In the 1990s, typically 70 per cent of never-married women who married had cohabited with their husbands, compared with 58 per cent of those marrying between 1985 and 1988, 33 per cent marrying between 1975 and 1979, and 6 per cent marrying between 1965 and 1969. The proportion of spinsters who were cohabiting more than trebled between 1979 and 1993, from 7.5 per cent to 23.5 per cent. Additionally, in 1993, 25 per cent of divorced women were in cohabiting unions. In the early 1980s, half of Hrsl partnerships consisted of people who had married without cohabiting, about a quarter of people who were marrying after cohabiting, and just under one in five were cohabitants. By the early 1990s, over half consisted of people marrying aflcr cohabiling, and the rest were equally divided between those who had married without having cohabited, and cohabitants (Haskcy, 1999). The peak age for women to cohabit is 20-24 years; 20 per cent of women in litis age group were cohabiting in 1995-6. Cohabitations have tended lo be shorl-lived The decline of marriage and the rise of cohabitation 35 (under two years) and childless, but during the 1980s children were increasingly being born within these unions. The changing patterns of birth registration provide some evidence of this (Table 2.5). By 1994, 58 per cení oťbirlhs were registered by couples living at the same address; teenage unmarried mothers are the least likely to be cohabiting. Table 2.5 Registration of births outside marriage, England and Wales, 1964-98(%) 1964 1971 1981 1994 1997 1998 60 54.5 41.8 22.7 21.3 20.8 40 45.5 58.2 77.3 78.9 79.2 7 8.4 12.8 32.4 37.0 37.8 Source: OPCS, Birth Statistics, Historical Series I837-19S3. series FM1 wo. 13 (London: HMSO 1987) Tables 1.1 and 3.7; ONS, Birth Statistics, series HM I, no. 27 (London: Government Statistical Service 1999) Tables 3.9, 3.1. Increases in divorce, cohabitation and child-bearing outside marriage have all contributed to the separation of marriage and parenthood. This is a very different phenomenon from the separation of sex and marriage that was forecast in the 1920s (Joad, 1946; Russell, 1985) and that was actually observable by the 1960s. Between 1970 and 1990, the percentage of lone-mother families more than doubled (Table 2.6). This was due in large measure to the rise in the divorce rate, which increased almost threefold over the same period, but also, since 1985, to the increase in unmarried motherhood. When cohabiting relationships with children break down, the women usually enter the statistics on lone motherhood as unmarried mothers. In other words, the rise in lone motherhood is to a significant extent the product of a process of 're-labelling' (Kicrnanerfl/., 1998). Figures on cohabiting relationships with children have only recently become available. Data from the British Household Panel Survey (BHPS), which has followed a sample of 10 000 adults annually since 1991, have shown that, among women born between 1950 and 1962, 9 per cent of those cohabiting had a baby; while the figure for those born since 1963 is 18 per cent. Kiernan's (1999) analysis using 1992 BHPS data showed 17 per cent of women aged 25-29 having their first child to be in a cohabiting relationship (a figure that is still low compared to the 53 per cent for Sweden or 46 per cent for France). BHPS data project that of every 20 cohabiting couples, 11 will marry, eight will separate and one will remain intact and unmarried after 10 years (Gershuny Sole registration Joint registration Births outside marriage as a percentage of all births 36 The end of marriage? and Berthoud, 1997). Table 2.7 shows that cohabiting couples were the fastest growing group in the 1990s, bat only 4 per cent of families are cohabiting Tahle 2.6 Distribution of lone-mother families with dependent children according to marital, status, Great Britain, 1971-98 {% of all families with dependent children) 1971 1974 1981 1984 1991 1998 Single lone mothers 1.2 1.0 2.3 3.0 6.4 9.0 Separated lone mothers 2.5 2.0 2.3 2.0 3.6 5.0 Divorced lone mothers 1.9 2.0 4.4 6.0 6.3 8.0 Widowed lone mothers 1.9 2.0 1.7 1.0 1.2 1.0 All lone mothers 7.5 10.0 10.7 12.0 17.5 23.0 Note: Estimates arc based on Uiree-year averages. Sources: J. Haskey, (1993) 'Trends in the Numbers of One-Parent Families in Great Britain'. Population Trends no. 71: 26-33; ONS, Living in Britain 1998 (London: Office of National Statistics, © Crown Copyright 2000) Table 3.4. Table 2.7 Families by type, UK, 1990-1 and 1995-6 (%) 1990-1 1995-6 Married couples With dependent children 44 41 With non-dependent children only 11 9 With no children 22 23 All married couples 77 73 Cohabiting couples With dependent children 3 4 With non-dependent children only — -— With no children 5 7 All cohabiting couples 8 11 Lone parents With dependent children 12 13 With non-dependent children only 4 3 All lone parents 15 16 All families 100 100 Source: P. Newman and A. Smith, Social Focus on Families (London: Office of National Statistics, © Crown Copyright 1997), p. 11. The decline of marriage and the rise of cohabitation 37 couples with children. General Household Survey data show thai, in 1998, 10 per cent of all families with dependent children were cohabiting couples.' One of the main issues in respect of all the statistics on family change has been the extent to which they signal a decline in obligation and cotnmilment and a concomitant rise in selfish individualism. Are people just more willing to walk out of relationships, even where there are children, on what earlier generations might have judged to be little more than a whim? The rise in the cohabitation rates has caused particular concern because it has been the driver of change in the late twentieth century and because cohabitation appears to offer a relationship with minimal ties. Mansfield and Collard (1988) described marriage as a 'strategy for the rest of your life', while early research on cohabitation showed that cohabitants saw fewer barriers to ending iheir relationships (Newcomb, 1981). Rindfuss and VandenHeuvel (1990) have argued that cohabitation is not an alternative form to marriage. Rather, cohabitants more closely resemble never-married people. Newcomb (1981) identified an element of 'Linus-blanket cohabitation' among young cohabitants, indicating both the need for emotional security and dependence. However, Beck and Beck Gernshcim (1995) have suggested that similar needs increasingly characterise marriage. This may be particularly Ihe case with young marriage, something that is much more common in the UK lhan the rest of Europe. However, the data indicate that cohabitation is more unstable lhan marriage, four times niore so according to British Household Panel Dala (Gcrshuny and Bcrthoud, 1997; Ermisch andFrancesconi, 1998). Maclean and Eekelaar (1997) found that 80 per cent of the children of the cohabitants in their sample who parted were under five years old, which compares with 31 per cent of the children of divorcing couples (Newman and Smith, 1997). Axinn and Thornton (1992) concluded that cohabitation selects the kind of people who are more prone to break-up. However, others have argued thai differences in values rather than characteristics explain the greater likelihood that cohabitants will dissolve their relationships (Clarkberg et al., 1995; Nock, 1995b; Lye and Waldron, 1997). Cohabitants are reported by many to have more egalitarian attitudes and to value independence more highly, although the meaning of independence differs, being associated with wanting a career, control over income and less commitment to sexual exclusivity by some, and with a more diffuse desire for autonomy and s elf-fulfil men t by others (Lye and Waldron, 1997). Newcomb (1987) and de Singly (1996) have highlighted the way in which cohabitation permits both the quest for independence and relatedncss. In Drew's (1984) analysis of pre-marital cohabitation, couples were found to be committed to individual aims which they hoped to connect to ones that could be shared together. Askham's (1984) work on marriage identified a very similar tension between independence and relatedness, but the tendency in the literature is to 38 The end of marriage? assume that the balance is tipped in favour of the former in cohabitation and in favour of the latter in marriage. The concern that cohabitation may be increasing at the expense of marriage has been heightened by concern that it appears to be relatively impermanent and a less 'committed' form of relationship, with higher rates of dissolution. Indeed, a Canadian study that attempted to operational ise Giddens' concept of the 'pure relationship', in which the partners are committed only for as long as they feel that they personally benefit, concluded that cohabitants came closest to matching the criteria developed (Hall, 1996). It is tempting to conclude that the process of 'individualisation', which according to Beck and Beck Gemsheim (Í995) pulls men and women apart, but at the same time makes a close relationship attractive, would favour cohabitation over marriage. Nevertheless, there are many different forms of cohabitation and the reasons for each may be very different (McRae, 1997a). In addition, interpretations of the evidence are in any case contested. Pre-marital cohabitation has long been justified as a form of trial marriage. Gillis (1986) has suggested that it resembles older patterns of bethrolhal. Certainly the average duration of pre-marital cohabitation is very close to the traditional period of engagement. Cohabitants were asked by the British Household Panel Survey in 1998 what they expected the outcome of their relationship to be, and 70 per cent answered 'marriage', although in fact only 60 per cent go on to marry. American research has indicated that the effect of pre-marital cohabitation on marriage may actually be negative (Thomson and Colella, 1992). Indeed Cheriin (1992), who is considerably less pessimistic regarding family change than many of his US contemporaries, has argued that cohabitation is a relationship that the parties believe should be ended if it fails to provide satisfaction, and that people take these attitudes into marriage: Cohabitation comes with the ethic that a relationship .should be ended if either partner is dissatisfied, this after all is part of the reason why people live together rather than marrying. Consequently the spread of cohabitation involves the spread of an individualistic outlook on intimate relations, (pp. 15-16) Rindfus and VandenHeuvel (1990) found that among their young cohabitants, cohabitation was seen as a way of securing intimacy without making any long-term commitment. The findings of Schoen (1992) and Clarkberg et al. (1995) were similar, and stressed the liberal values of cohabitants and the way in which these were indicators of preferences for a type of relationship that was essentially different from marriage. However, Brown and Booth (1996) found on the basis of their national sample survey of US households and families that the relationship quality of cohabitants with children was similar to their married counterparts; it was the The decline of marriage and the rise of cohabitation 39 cohabitants who did not intend to marry who were different. Kiernan and Estaugh's (1993) British data showed that the attitudes of cohabitants with children were less liberal than those of childless cohabitants and were very similar to those of married parents, The most recent European research by Kieman (1999a) provides convincing evidence of the similarity between first partnerships that are maniages and those that begin as cohabitations but convert into marriages. Kieman found very little difference in most European countries between these two groups in terms of the proportion that remained intact after 10 years, which raises a question mark over Cherlin's idea that cohabitation necessarily affects people's approach to marriage. However, the dissolution rate is much higher for first partnership cohabitants who do not marry, especially in Britain. Furthermore, cohabiLanls with children are less likely to convert into married partnerships and more likely to dissolve. Half of first-time cohabiting parents dissolve their relationships by the time their child is five. All this serves to show that the population of cohabitants must be disaggregated (Manting, 1996). Some cohabitants may avoid entering a legally constructed relationship in order to make it easier to move on if it does not come up to their high expectations. Some may make a principled decision about cohabitation being more suited to a relationship in which both contribute equally. The idea of the pure relationship is premised upon a degree of equality between the partners, and assumes a large measure of material well-being. Smart and Stevens (1999) have referred to those making a principled decision about cohabitation as 'reflexive'. However, others may be responding to difficult circumstances. Living together is popular among both students and the young unemployed. But cohabitants who do not marry are more likely to have lower levels of education, to have no religious affiliation and to have experienced the divorce or separation of their parents (Kiernan, 1999b). McRae (1993) investigated 228 mothers who had cohabited either before or in place of marriage and 100 never-cohabiting, married mothers. She concluded that cohabitation represented a 'rational response' to low male wages and economic insecurity (see also Ermisch and Francesconi, 1998). Marriage is 'practised most often by those with something to transact' (McRae, 1993, p. 106). In this analysis, it is material circumstances rather than values that are most important in explaining this form of cohabitation. Smart and Stevens (2000) have called this low-income group the (rational) 'risk' takers. BHPS data have shown that among this group a rise in the man's wages means that the couple is more likely to decide to convert the cohabitation into a marriage (see also Smock and Manning, 1997, using US data). In fact the majority of UK cohabitants with children are disproportionately ill-educated, young and poor. In the 1960s, this group may well have had shot-gun weddings, quite possibly divorcing later. Given the rates of cohabitation breakdown, it is now likely that these women become lone mothers by a different route. 40 The end of marriage? It is tempting to write of the 'rise and decline', if not the 'end', of marriage in the twentieth century, with marriage becoming virtually universal in the immediate post-war decades and seemingly becoming much less popular in the closing years of the century. As Smock (2000) has remarked, the key question now seems to be not who cohabits, but who does not. Even remarriage rates, which were high in the 1970s, have fallen dramatically (Table 2.8). The statistics relating to the higher rates of dissolution for cohabitants who do not marry have been taken to signal a greater degree of individualism. In the view of Newcornb (1981), cohabitation meets the desire for individualism and intimacy, but it may just as plausibly be argued that it meets the needs of late twentieth-century young people who became sexually active earlier and remain economically dependent for longer (de Singly, 1996); of young people who are sexually active with no property and few skills; or of women who experience conflict between a desire for autonomy and the sanctions of marriage (Adams, 1998). Tabic 2.8 Remarriage numbers and rates by sex, England and Wales, 1961-97 Ail ages Rate per 1000 Total number of No. (000s) divorced people decrees made aged 16 and over absolute (000s)a Men 1961 1971 1981 1991 1995 1997 Women 1961 1971 1981 1991 1995 1997 Note: s Each decree absolute creates a divorced man and a divorced woman. Sources: ONS, Population Trends no. 88, (1977), Tables 15 and 22; Population Trends no. 47, (1987), Table 15; ONS, Marriage, Divorce and Adoption Statistics 1997, Scries FM2 no. 25 (London: Office of National Statistics, © Crown Copyright, 1999) Tables 3.7, 3.8, 3.12 and 4.5. 18.8 42.4 79.1 74.9 77.0 76.8 18.0 39.6 75.1 73.4 76.9 77.1 162.9 227.3 129.5 61.6 56.1 47.9 97.1 134.0 90.7 49.0 45.4 41.0 25.4 74.4 145.7 158.7 155.5 146.7 25.4 74.4 145.7 158.7 155.5 146.7 The decline of marriage and die rise of cohabitation 41 To some extent, all these explanations see the emergence o ľ cohabitation as a rational solution to complicated changes in beliefs and behaviour. 11 is not possible to make simple links between less marriage, more cohabitation and more individualism. Cohabitation is now both alternative and sequential lo marriage. Haskey's (1995, 1999) analysis shows thai patterns of marriage, divorce, singlehood and cohabitation are becoming much more complicated within the life-course of the individual. Periods of cohabitation may precede marriage and follow divorce. Whereas in the nineteenth century marriage rates were a reasonable proxy for the employment rate, this was no longer so in the late twentieth century. Belter wages together with a national minimum level of welfare secured by the state have ensured that people can adopt the kind of post-materialist values identified by Inglehart (1997). Following the ideas of Beck and Beck Gernsheim, Kuijsten (1996) concluded that there is 'convergence towards diversity' as individuals construct new biographical models involving serial cohabitations, marriages and divorces. Prinz (1995) developed a model of cohabitation with four stages, in which the fourth and final stage is cohabitation with children. He suggested that at this point cohabitation looks like marriage, but largely because marriage has become more like cohabitation. However, such a staged model is tidier than the reality. Indeed, Gillis (1997) has turned the argument around, suggesting that all intimate relationships now resemble conjugal ones. 'The perfect couple', loving and committed, has become the standard for pre-marital and cohabiting relationships, This may be the case at the level of expectations, but the important point is that the meaning of cohabitation varies, just as does that of marriage. In these circumstances, marriage becomes less central as a unifying cultural experience (Nock, 1995b). The issue remains as to whether marriage, as an institution as well as a relationship, is in some way different and whether it matters. Marriage is a public institution and requires a public declaration of commitment. Schoen and Weinick (1993) have commented on the fact that cohabitation is more informal and that cohabitants emphasise achieved statuses - in the form of education and employment, for example - rather than ascribed (family) characteristics. Cohabitation is almost certainly a more 'private' arrangement, but whether it amounts to an 'incomplete institution', lacking agreed standards and relationships with kin (Nock, 1995) is another matter. It may be argued that it is increasingly an issue as to why people marry at all, given that cohabitation has become normal practice. The emergence of widespread cohabitation raises large issues for family law and social policies. Increasingly, the separation of marriage and parenthood has been reflected in legal changes; for example, all biological fathers were given the obligation to maintain their children by Ihe 1991 Child Supporl Aci. But there is no general body of law relating to cohabitation in Ihe UK. 42 The end of marriage? Sometimes cohabitants are treated as if they are married, as in the case of means-tested benefits, and sometimes as if they are not, as in the case of insurance-based benefits (Harris, 1996). Family change and in particular the decline in marriage rates, marital breakdown and the emergence of cohabitation, has undermined all manner of legislation that assumed the existence of two-parent, married-couple families and, to a varying degree, the operation of the male breadwinner model. Furthermore; government has to decide whether to codify cohabitation. Interestingly the criteria developed by the social security system for establishing the existence of a cohabitation consisted of public manifestations, for example, a change of name, or taking holidays together. In the 1990s, the Solicitors Family Law Association and the Law Commission (Rodgers, 1999; Gourict, 2000) have discussed the idea of defining a cohabitation for legal purposes, the latter leaning towards public manifestations and the former towards evidence of personal and financial commitment to the relationship; however, both agreed that the amount of time that it has lasted should be a criterion. From the legal perspective, Brenda Hoggetf's (1980) observation almost 20 years ago that the distinction between marriage and non-marriage would become relevant mainly to the childless couple where one partner was inactive in the labour market was prescient. But law-makers (and academics) have great difficulty in working wiLh the grain of family change, and the inclination is often to try and resuscitate older marital forms, which is all the more difficult now that the prescriptive frameworks supporting them Have been eroded. Marriage has changed too. It remains a public institution, but it is very more open (to negotiation) in terms of the nature of the relationship between the partners than it was. The expectations people hold of marital relationships and of intimate relationships in general are no longer as sure as they once were. Thus it is not surprising that both negotiation between the partners and a reluctance to make judgements about the behaviour of other couples has increased. NOTE I. i am grateful to Kathleen Kiernan for this information (personal communication). PART II