The processes by which social inequality is transmitted from one generation to the next are important for all societies. Notions of how ‘open’ a society is strongly legitimize or delegitimize its social and political order. In a fully open society each individual will achieve as much as his or her talent allows, without being stopped short by ‘inherited’ inequalities, most notably social background. Societies open to ‘talent’ are better geared for innovation and economic growth, and we are inclined to think of them as fairer than societies that block the social ascent of their talented members in favour of inherited positions. Fully open or meritocratic societies do not exist, but some societies and some periods are more open than others. Has there been a trend towards more social mobility in all or some parts of the world? That is a fundamental question for sociology and social history. Many historians and sociologists think that there is a trend away from traditional societies, where a person's social position was essentially inherited, to more open ones, where social position depends upon individual achievement. Is there a convergence in levels of mobility between parts of the world? If so, where and when did those trends start? And what were the determinants of social mobility in the various regions and historical periods?
Variations in mobility patterns have long since interested historians and sociologists, from de Tocqueville, Marx, Weber, Sombart, Sorokin and the New Urban Historians to the flourishing community of stratification sociologists today. When de Tocqueville travelled in the USA early in the nineteenth century he noted an exceptionally high degree of mobility:
There is still a class of menials and a class of masters but these classes are not always composed of the same individuals, still less of the same families; and those who command are not more secure of perpetuity than those who obey … At any moment a servant may become a master, and he aspires to rise to that condition; the servant is therefore not a different man from the master.Footnote 1
The contrast with Europe, in his case France, was stark, and similar remarks were made by others, including John Stuart Mill,Footnote 2 Marx,Footnote 3 and Sombart.Footnote 4 As John Stuart Mill phrased it, in relation to England:
So complete, indeed, has hitherto been the separation, so strongly marked the line of demarcation, between the different grades of labourers, as to be almost equivalent to an hereditary distinction of caste; each employment being chiefly recruited from the children of those already employed in it, or in employments of the same rank with it in social estimation.Footnote 5
The question has received so much attention because the stakes are so high. It is not only a core question of history, sociology and economics, but also a subject of great social relevance. High rates of social mobility are seen as proof of the absence of institutional barriers to each and every individual's use of his or her talents. If that is true, then the resulting degree of social inequality is justified even if it is extremely high. As the Chairman of the US Federal Reserve, Ben Bernanke, put it recently in defence of this high degree of inequality: ‘A bedrock American principle is the idea that all individuals should have the opportunity to succeed on the basis of their own efforts, skill and ingenuity.’ Indeed, faith in meritocracy is seen as the ultimate justification for high levels of inequality:
Although we Americans strive to provide equality of economic opportunity, we do not guarantee equality of economic outcomes, nor should we. Indeed, without the possibility of unequal outcomes tied to differences in effort and skill, the economic incentive for productive behavior would be eliminated, and our market-based economy – which encourages productive activity primarily through the promise of financial reward – would function far less effectively.'Footnote 6
This is not a belief of only the high and mighty. A strikingly large proportion of the American population for instance believes that they stand a good chance of becoming upwardly mobile and rich.Footnote 7 In short, there is a widely shared opinion – at its peak probably in the USA but in much of the rest of the world too – that a meritocracy justifies both high rates of social mobility and a great degree of social inequality. If the claim is correct, historical, sociological and economic research will demonstrate that few of the institutional barriers to social mobility in past societies, which will be discussed later, are still in place.Footnote 8
Here I present a short synthesis of the historical and sociological literature that has proposed explanations for differences in social mobility over the past three centuries. It tries to place the subsequent articles in this special issue of Continuity and Change in a context by discussing three forms of social mobility: intergenerational mobility (between parent and child); social homogamy and heterogamy, or mobility at marriage (measured mostly between the father of a bride and the father of her groom); and then career mobility (taken over the course of a whole life). These forms of social mobility are not identical and indeed they fully merit being objects of study on their own, but they share some communalities with regard to the forces driving them. I will pay most attention to intergenerational mobility in history, for there is no convenient recent summary of the vast literature on its determinants, unlike in the case of social homogamy and the history of the career. (See the surveys on social homogamy and the history of the career mentioned later.)
INTERGENERATIONAL SOCIAL MOBILITY
The intellectual origins of this field of research as we know it today can be said to be modernization theoryFootnote 9 and Blau and Duncan's status attainment model.Footnote 10 The latter predicts the class of destination of a son, based only on the class of origin of his father and on the education of both parent and child. The model has been enormously successful and is still widely used today. In later versions the range of personal characteristics influencing social mobility has been considerably broadened. Modernization theory predicted that as societies industrialized or ‘modernized’, employers would increasingly recruit their personnel by reference to the individual's merits or achievements rather than to that individual's parents (ascription). In the absence of countervailing tendencies, that leads to a loosening of direct bonds between class of origin and of destination, a process from ascription to achievement. In so simple and universal a form, however, modernization theory is moribund. Historical contexts have to be taken into account.
The earliest studies on intergenerational social mobility focused solely on observed total mobility rates, with the number of sons who are in a social class different from that of their fathers expressed as a percentage of the total number of sons.Footnote 11 It soon appeared that observed total mobility rates from the 1950s onwards differed both among and within countries and did not vary systematically with the level of industrialization. With that knowledge, the scholarly debate has shifted to variations in relative mobility, also termed ‘openness’ or ‘fluidity’. Openness is mobility that is not due to changes in occupational structure. Erikson and Goldthorpe concluded that meaningful variations in relative mobility between countries are absent in the survey material, as are trends.Footnote 12 Ganzeboom, Luijkx and Treiman, however, concluded that relative mobility grows by 1 per cent per year.Footnote 13 If that is so, it will require a long time horizon to observe such slow changes, and this can be done best by making use of long-term historical data. Reviews of prior work make clear that no consensus exists about the development of social mobility over time, nor about the driving factors between temporal and regional variation.Footnote 14 To some extent that has to do with disciplinary boundaries,Footnote 15 but it is also due to the rather small time horizon of both historical and sociological studies, and the fact that institutional contexts have not been sufficiently taken into consideration.Footnote 16
The literature shows a great variety of competing claims for determinants of social mobility.Footnote 17 All such determinants, derived from conflicting theories, might well have been relevant at different times and places, but at the moment we do not know their relative importance. The single most important thing we have learned from the historiography of social mobility published over the past few decades, however, is that not only do the personal characteristics of those who end up in a mobility table matter but so too do the core institutions in the societies they lived in, and that includes not just economic institutions such as guilds or industrializing factories but trade unions too, as well as the political structure and schools, for example.
I shall begin, however, by discussing the more salient personal characteristics which have a bearing on social mobility patterns and processes and then go on to discuss institutional contexts. In the literature on determinants of social mobility, the personal characteristics that are thought to influence social mobility most are gender, parents, siblings, education and migration.
In pre-industrial societies girls were less likely than boys to work in the labour market, and if they did so the range of occupations open to them was more limited, and this is so even if we take into consideration that there is often a significant under-recording of female occupations in historical sources. Parents tended to invest more in the education of boys, for the rewards of male education and training were higher than those for girls. If parents in their old age became dependent on support from their children, it was also more profitable to seek to raise the earning power of boys. There was more volatility in the choice of occupation for daughters than for sons. French and German autobiographies from the period of industrialization ‘all suggest a certain contingency that governed occupational choice in early adolescence. For girls especially, there seems to have been little conscious planning, little ability to imagine occupational futures.’Footnote 18 Sons have generally been more ‘privileged’ than girls, the more so if the legal system (primogeniture) and cultural norms dictated it, giving of course the greatest advantage to first-born boys over their younger brothers. Over time one would expect the social openness of women to have increased more than that of men, as more women entered the labour market, as modern bureaucratic labour markets came into existence and as social norms and laws (including marriage bars, inheritance laws and customs favouring the eldest male child) grew less prevalent. Furthermore, the growth of social security schemes made parents in their old age less dependent on their children, reducing the necessity to ensure that at least one of the boys earned enough to help his aged parents, while at the same time still leaving some parents dependent on the care of a child, notably a daughter.
Parents can influence the choice of occupation of their children, by helping them to find a job. As well as having the right connections and financial resources, parents have cultural resources; they can help their children with reading and writing, familiarize them with a certain type of work, and mould their children's aspirations. Over the last decade the effect of mothers on the schooling and later the work of their children has become a research topic in its own right.Footnote 19 It appears that in contemporary societies a mother's occupation has an effect independent of that of the father. Miles argues that fathers are helpful in getting their sons into a job, acting as ‘occupational brokers’, but for girls their mothers can familiarize them with a certain type of work, and use social contacts in helping the daughter to get a job.Footnote 20 Certainly for servants it has been shown that a mother's occupation as a servant tended to encourage a girl to become one herself.Footnote 21 Mothers with specific skills or property, such as midwives or shopkeepers, could also transfer such livings to their daughters. For girls a ‘Cinderella effect’ has been signalled in the literature:Footnote 22 in the past, if a girl's mother died her father often remarried quite soon and in some such instances a girl might have wanted to get out of the house as soon as she could, even if to do so meant rushing into an inappropriate job. Thus a remarrying father might mean downward mobility for his daughter.
The number of siblings and one's place in the pecking order within the family can influence social destination. A large number of siblings means that parents, by a process known as ‘resource dilution’, have to divide both their dedication and their income among many children, although they can of course favour any particular one. In some societies polygamy is practised, increasing the number of siblings; that might lead to further ‘resource dilution’, or perhaps to the children of the first wife being favoured. Small families may have been more successful in helping their children to reach higher positions, especially in the case of small families embracing the novel techniques of family limitation.Footnote 23
Apart from siblings and parents, the resources of grandparents are relevant too.Footnote 24 Note that this relates to the impact of grandparents over and above their indirect impact through their own child, who is the mother or father of the subject. It is generally held that such multigenerational influences are weak, but in certain cases they might make a difference, for example if a father died young, or a grandfather possessed a certain attribute – a title or reputation – that he had not passed on to his child, as was sometimes the case among grandfathers in Imperial China.Footnote 25
Education is important as it furnishes skills needed for particular jobs. Not only one's own education but that of one's parents can be important because well-educated parents choose better schools, help with homework and encourage schooling. As literacy in general increases, the effect of illiteracy might tend to become greater: those who could not sign their names for themselves in twentieth-century Europe became the truly disadvantaged.Footnote 26
Being a migrant can influence one's prospects of finding a job. There are several rival explanations for this.Footnote 27 Migrants might be a positive selection from the population in general (‘the ambitious move on’) or they could be a negative selection – ‘a floating proletariat’.Footnote 28 In moving to a new city they might deprive themselves of help from friends in finding a job, especially if they belong to a disadvantaged ethnic group, but they might profit from opportunities in a new labour market which matches the skills they can offer (an argument made by de Tocqueville and others subsequently to explain the assumed greater openness of the USA, where new frontiers opened up new possibilities to climb).
The effect of personal characteristics on social mobility depends at least partly on the institutional context: some societies discriminate against girls while others stimulate their occupational rise by positive action; whether the skills a migrant has learned prove useful depends on the nature of those skills and also on demand in the receiving society.Footnote 29 Institutional settings are crucial. Failure to take contexts into account might be an important explanation of the often contradictory results that have been found for determinants of mobility. The relevant institutional contexts are: industrialization and other economic changes; educational expansion; guilds, trade unions and other professional organizations; political regimes; urbanization; wars; marriage bars and inheritance patterns.
Industrialization has long been the institutional context attracting most of the attention of stratification sociologists. Modernization theorists believe there is a global historical trend towards social ‘openness’.Footnote 30 Modern societies, it is argued, are both predicated upon and characterized by their openness. The utilization of talent and the expression of preferences are free from the inhibiting forces of tradition and power differences. ‘A competitive industrial system … will increase social mobility, raising the gifted, and lucky, and lowering the inept, lazy, and ill-fortuned. This is the kind of thing one sees in eighteenth or nineteenth century Germany … or in nineteenth and twentieth century Japan … or in France … or in the India of today’.Footnote 31 Although the theoretical meaning of ‘modernization’ is quite broad in the works of some authors,Footnote 32 it was often operationalized in rudimentary form as the number of steam engines or per capita energy consumption, but that is a narrowing caused by the relatively easy availability of such indicators and, while understandable, it is unfortunate. Recent work on the history of the career also supports the existence of a relationship between economic change and social mobility,Footnote 33 but does not look towards industrialization, pointing instead to processes of economic specialization and the rise of large bureaucracies with their own internal labour regimes, such as railway companies or banks, which select personnel through advertisements and open applications.
Another institutional context to attract attention for its influence on social mobility is systems of education and their development, often called ‘educational expansion’. Individuals have varying degrees of schooling, and societies differ in their educational systems. Economists stress investment by parents in human capital, as well as the progressive nature of public investment in human capital, in other words the degree to which children from poor families are subsidized to go to school.Footnote 34 In theory, then, the more parents or the State invest in education, the more mobility there is. And indeed, in a meritocratic society, a son or daughter obtains a job suited to his or her skill level, and skill levels are learned in schools.
However, the historical record might not invariably fit this straightforward view. There is a competing opinion, one that stresses the importance of education too, but more in the form of certificates and diplomas which serve, rightly or wrongly, as shorthand for skill. By ‘wrongly’ is meant that, as social reproduction theory stresses, elites use certificates to circumvent meritocracy. Collins writes that ‘the demands of any occupational position are not fixed, but represent whatever behaviour is settled upon in bargaining between the persons who fill positions and those who attempt to control them’,Footnote 35 And further that ‘education is an artificial device for monopolizing access to lucrative positions’.Footnote 36
Elites have the funds to pay the cost of their children's attendance at ‘good’ schools – elites who do not need income from child labour and who enjoy a way of life that better prepares children for these ‘good’ schools, for example through motivation or simply by helping with homework. Social reproduction theorists thus differ fundamentally from meritocratic theorists with regard to the role of schooling: schooling reinforces and tightens class bonds through the generations, rather than loosening them. Elites may have made more use of credentials over the years precisely because the elites' direct influence has eroded. Because it has become increasingly difficult to place a child directly into a privileged position, an indirect ‘compensation strategy’ of social reproduction is used.Footnote 37 It is not easy to distinguish between credentialist and true meritocratic societies. Treiman, Ganzeboom and Rijken formulate social reproduction theory in terms of trends over time:
While … equality of educational opportunity may be increasing at low levels of education, access to universities and other elite institutions is seen as monopolized by the rich and the powerful. Thus the claim is that the effect of social origins on educational attainment is greatest at the high end of the educational distribution. From this, it follows that, ceteris paribus, as education expands the dependence on social origins should increase.Footnote 38
They propose to deconstruct the effect of the expansion of education on social mobility within a society into a function of the level of education in that society – where the more years of schooling a population has, the more open it will be – and into an effect of the unequal social attendance of education: if educational inequality is higher, social closure will be greater.
Credentialism is manifested too in the influence of guilds, professional organizations and trade unions. In a meritocratic view of the world, such institutions impart skills and certify to an employer that a person has what skills are needed for a certain job. But they may also be seen as institutional hindrances to newcomers. Guilds were most open to the sons of their masters, and in any case generally open to neither women nor Jews. Trade unions, it is often said, do most to further the interests of their members, predominantly skilled male workers. Penn reports that in England during the Great War a shortage of men forced both trade unions and employers to allow women to do skilled work previously forbidden to them because they were supposedly under-qualified, but despite this lack of formal qualifications women managed well during the War.Footnote 39 After it, they were seen once more as unqualified and their jobs were given back to returning servicemen. A similar situation occurred after World War II.Footnote 40
Political regimes matter in loosening or tightening the occupational bonds between parents and children. Prime examples in the literature are communist regimes.Footnote 41 They set out to block the transmission of top positions to the children of the bourgeoisieFootnote 42 while at the same time recruiting the ‘socialist intelligentsia’ from the rank and file.Footnote 43 They intended to provide equal opportunities through affirmative action. Other ways in which they tried to do that were by favouring students from backward regionsFootnote 44 or by confiscating private property, which prevented people passing assets on to the next generation.Footnote 45 Income differences in communist societies narrowed. Furthermore, communist regimes in theory favoured equal opportunities for men and women, which was sometimes expressed in laws or quota systems.Footnote 46 Rijken reports that the gender gap in opportunities was indeed lower under communist regimes, and that the effects of ascription were lower.Footnote 47 That need not be so, either because good intentions may be ineffective or because, behind the façade of equality propaganda, communist elites, like other elites, might be expected to try to pass their own privileged positions on to their children.Footnote 48 Some studies state that communist regimes begin by increasing occupational mobility but, after a generation or so, revert to social reproduction – of the new communist regimes, in their case.Footnote 49
Political regimes can be divided into groups other than those defined by the dichotomy between communist and non-communist regimes.Footnote 50 In a milder form, the same theoretical arguments favouring increasing social fluidity can be made for social-democratic regimes, and perhaps in practice even more so because such regimes come to power in free elections and so are unable to ‘hive off’ their positions of privilege or hand them over to their children collectively. States are also major employers, so they are able to influence occupational achievements, notably by some sort of affirmative action for groups that are underprivileged or discriminated against. So political regimes with large public service sectors can choose to have a positive effect on openness.
The list of institutional contexts influencing intergenerational social mobility can be expanded to include other contexts, notably urbanization, wars, marriage bars and inheritance patterns. Cities are often believed to loosen the bonds between parental occupations and those of their children in several ways. Schooling as well as transport and communication facilities are generally more developed in cities, where large bureaucratic institutions with formal recruitment procedures first appeared. Moreover, cities offer children more scope to escape from parental pressure, especially if they move there from rural areas.
Wars and revolutions are social eruptions which sometimes have the power to change established patterns of occupational choice. For men, they have often meant geographical displacement and a fresh look at society, including one's own place in it. Furthermore, joining the army could be a way for poorer children to obtain vocational training, a way out of their social class. Since times of war drain the labour market of young men, such times also offer a window of opportunity, even if only a small one, for women to do work previously done by men. So we expect wars to increase social fluidity, at least in the short term, and to do so more effectively for women than for men. Historically speaking, it is quite clear that normative and legal restrictions made the choice of occupations smaller for women than for men.Footnote 51 The existence of marriage bars meant that women had to stop working after marriage, and that certainly discouraged educational investment. Finally, inheritance systems might favour the eldest son to the detriment of any other children, as with either de jure or de facto primogeniture.
CAREER MOBILITY
Given the relatively large number of studies that have been written about intergenerational mobility – meaning changes of occupation from parent to child – one might expect a relatively large number of studies on the changes of occupation of one person over his or her life course, referred to as career mobility (as opposed to changes of occupation between the generations). After all, both relate to occupational change occurring at the same historical time, and both will be influenced by a combination of the characteristics of the persons in question and contextual influences, and in many cases the same personal and contextual determinants. But such an expectation would be false. The number of studies on career mobility, past and present, is clearly lower than that on intergenerational mobility and fewer than the number of articles and monographs devoted to social homogamy. In part that might be a reflection of the difficulties faced by sociologists studying contemporary societies, and by historians studying past populations, in finding data able to capture occupational change among the same set of individuals over their life course. Whereas historians may use one single marriage certificate from an abundant amount available to them to infer both the marital and intergenerational mobility of a person, for career mobility such a certificate is of limited use. To study career mobility for the population at large, or at least a sizeable share of it, one needs to link marriage and other vital registers, or to link censuses, or to use one or more population registers and follow the persons in question all the way through them. Either method requires far more effort than is usually needed to obtain the data for either intergenerational mobility or social homogamy.
This unfortunate situation has led not only to a relative shortage of data and empirical historical studies based on them, but also perhaps to a paucity of theoretical claims regarding the driving forces that have shaped careers or that have determined changes over time. These claims have recently been summarized.Footnote 52 While there is little point in repeating those summaries here, it is worth noting that, just as in the history of intergenerational social mobility, it is possible to distinguish two conceptually different sets of determinants of changes of occupation over the life course: personal characteristics and institutional contexts.
Personal characteristics that influence career mobility include gender, parents, siblings, education and migration, just as in the case of intergenerational mobility, and for much the same reasons. To give an example relating to gender, until recently the number of different jobs open to women was much smaller than that for men, which was due partly to gender inequalities in education caused by different investments by parents for boys and girls, partly to the demands of child care, and partly due to contextual impediments to women in the form of policies, rules and regulations, and the laws of unions, firms and governments.Footnote 53 This example shows that, apart from personal characteristics, contexts have mattered in shaping careers differently between regions and periods: some contexts discriminate against girls, others (although not many, it must be said) stimulate their occupational rise by positive action.
As was the case with intergenerational social mobility, institutional settings are of prime importance. Due to the relative lack of historical studies of the career, it is not easy to be sure what constitutes the set of relevant institutional determinants. The recent summary by Brown, Mitch and van Leeuwen lists the following potential influences: industrialization, the rise of hierarchical bureaucratic management structures, the development of internal labour markets, the spread of education, migration regimes and discrimination by race, religion and sex.Footnote 54 Some of them are identical to the crucial institutional factors identified above as shaping intergenerational social mobility.
Industrialization can be expected to have stimulated social mobility in two ways. The first is in an economic sense, by changing the occupational structure, including the creation of new jobs needing to be filled by persons who were mobile – in the sense of having held a different job or coming from a different occupational origin. The second is in a social and cultural sense, by whetting the appetites of employees for occupational change and by stimulating a shift in the recruitment practices of employers, who needed to find the best man or woman for the new job.
The role of education appears to have been similar, certainly in the Western world. Over the last two centuries there has been a dramatic increase in the number of educational institutions providing, formally or informally, the qualifications needed for a new job. Those institutions provided universal primary schooling and an extension of secondary and tertiary education, but notable too was the rise of organizations of professionals to provide training, education, certification or licensing for chemists, notaries and medical practitioners, for example. In some ways they resemble the ill-fated guilds that also controlled entrance to the labour market.Footnote 55 Organizations of workers, that is unions, were active in trying to regulate the labour market in ways favourable to their members.Footnote 56
Employers were doing the same, faced as they were with the problem of employing and retaining qualified personnel wherever that was a positive requirement. The rise of hierarchical bureaucratic management structures and the development of internal labour markets from the late nineteenth century can be seen as institutional responses to the problem.Footnote 57 An internal labour market in a firm has its own set of rules and procedures to create a structure for the career paths of its employees through clearly defined job categories, wage scales, a personnel department and, perhaps most of all, the prospect of rising through the ranks in a clearly identifiable, more or less predefined way.
MOBILITY AT MARRIAGE: SOCIAL HOMOGAMY
The field of study of mobility at marriage is usually termed ‘social homogamy’, if one marries into one's own social class, or ‘social heterogamy’, if one marries into a different social class. In practice the terms are used interchangeably with those of ‘social endogamy’ and ‘exogamy’.Footnote 58 Whereas the historical study of the career is a recent idea and the number of studies is limited, social homogamy has a venerable research tradition, in ethnology, anthropology, sociology and history, as a recent survey shows.Footnote 59 More so than in the study of either intergenerational social mobility or the history of the career, where the influence of ethnology and anthropology is limited but that of economics more prominent, there has been an emphasis on cultural norms shaping marriage patterns by social class. Although it is not always phrased like that, there has often been an emphasis on institutional determinants. A now classic way to structure the determinants of social homogamy is to look at three clusters: individual preferences, third-party influences and the structural constraints imposed by the marriage market.Footnote 60 While the notion of individual characteristics influencing social mobility processes is probably somewhat broader than that of individual preferences, as will be seen shortly, it might still be true that in studies on social homogamy there has been a relatively greater stress on institutional determinants.
Individual preferences may in part be seen as random (a preference for ‘lively blue eyes’ over ‘sleepy brown eyes’, for example), or at least as usually beyond the domain of the historian. They may also refer to the wish to marry for love, a subject which has attracted a lot of discussion, following Shorter's claim that in this respect a revolution has occurred over the past two or three centuries – at least in the Western world.Footnote 61 Although romantic love can be seen as an individual preference, Shorter presents it as a universal drive of the human heart that had been routinely thwarted by parents, priests, mayors, neighbours and peers. When the authority of those forces began to be eroded and finally gave way, the human heart could follow its romantic desires. What can be seen as an individual preference can be seen too as a change in institutional contexts, due to migration and urbanization, mass communication, secularization and other factors. Another individual preference mentioned in the literature is that some, if not most, people prefer a partner who shares one's own values and tastes and cultural background, though they have no objections to marrying into an economically higher class.Footnote 62
Individual characteristics thought to influence inter- and intragenerational mobility have included gender, parents, siblings, education and migration, and so they differ from the individual preferences thought to shape marriage patterns by social class. While education and migration may in part be seen as the result of an individual preference (but also as a consequence of decisions by others, notably parents who might or might not invest in schooling or move to another place), gender, parents and siblings are not preferences but givens.Footnote 63 That is not to say that the literature on social homogamy has ignored those factors. On the contrary it is often acknowledged that parents, although they seldom formally arranged marriages in the Western world, tried to influence the choices made by their children. It has been acknowledged that having many brothers and sisters can be harmful to one's marital prospects, due to resource dilution, which has been argued too for intergenerational mobility, and it has been recognized as well that parents might have been inclined to place greater emphasis on finding a suitable marriage partner for a first-born son than for later-born sons or for daughters. Gender, parents, siblings, education and migration, however, are subsumed in the other two categories of the tripartite division of determinants of social homogamy: third-party influences and the structural constraints imposed by the marriage market.
The third parties affecting social homogamy are parents, peers and communal authorities, and their effect in the past is now generally seen as having been considerable, especially in agricultural regions. Segalen, for example, wrote about France:
In traditional society, problems that would today be considered personal, whether to do with the intimacies of the heart or of the body, were the responsibility of the community. The formation of the couple, as well as concerning the young people themselves, involved the two families and the entire social group … Family considerations weigh heavily on the individuals, who tend to disappear in the face of the wider aims of economic and social improvement of the family line. In these terms the couple is merely a link in the chain leading to the growth of patrimony or resisting the fragmentation of landholdings through inheritance. The individuality of the couple, or rather, its tendency towards individuality, is crushed by the family institution, and also by the social pressure exercised by the village community as a whole.Footnote 64
Segalen also described vividly how neighbours and peers could disapprove of certain marriage candidates, who might for example be too old or too young, or who perhaps came from another village. Much of the historical anthropological and ethnological literature in this respect incidentally deals with animosity and quarrels – vile looks at communal water places, kettle music in front of the houses of culprits, and fights – rather than describing gentle guidance towards a suitable marriage partner, probably because animosity and quarrelling have a greater chance of ending up in the historical record in an archive or of being remembered by aged people during interviews.
Institutional influences on the choice of a marriage partner are apparent too in the third cluster of determinants of social homogamy: the structural constraints of the marriage market.Footnote 65 This cluster can be split in turn into two subgroups: marriage horizons and meeting chances. ‘Marriage horizons’ are the geographical areas from which potential spouses were drawn, whereas ‘meeting chances’ refers to the likelihood of actually meeting a potential spouse within that area, and the factors determining that likelihood.
Nowadays potential ‘brides’ from exotic places can solicit marriage partners by sending love letters and descriptions of their background, their bodies and their intimate desires to millions of people unknown to them all over the world, or at least the Western world. A marriage is just one mouse click away, and if not a marriage then at least an exchange of emails or possibly a meeting or relationship. These unsolicited emails – and especially the more serious dating sites on the Internet – are a modern and no doubt valuable representation of a marriage market. The problem, if any, is adopting just a single choice amidst the abundance in the ether! In past societies the problem was not often one of abundance. Young sons and daughters of farmers and farmhands in particular were often faced with a rather limited supply of potential spouses, so that one might argue that the marriage horizon for villagers until well into the nineteenth century was generally limited to the distance they could walk in half a day at most, for they had to walk back as well in order to be home in good time.
While that is of course an oversimplification – because young men and women could decide to migrate or they might travel to fairs and young men could be forced to move around as part of their military service, for example – it does indicate that marriage horizons in past societies were for many people limited to an extent that is hard for us to imagine today, because in the course of the past two centuries horizons have widened greatly due to the growth of mass communication (love letters) and mass transport (air travel, railways, bicycles, trams and motor cars). Mass communication and mass transport are, then, two institutional factors which, as we have seen, have also been determinants of intergenerational social mobility.
A lot of attention has been given in studies of social homogamy to the factors influencing the likelihood of meeting – and eventually mating with – someone from another class within a given marriage horizon. Factors thought to influence this likelihood are: the rise of mass education, spatial social segregation, associational life, the rise of a modern labour market and the degree of religious or ethnic diversity. The role of education is in large part the same as discussed above: schools open up other social horizons because of the proximity of potential spouses from different social classes, while the curriculum itself might teach youngsters that they do not necessarily have to follow the opinions of their parents about what constitutes a good marriage. That is of course the more true if schools are mixed with regard to gender and social class, and it is especially true for higher education, such as universities or vocational training, as young people there are of marriageable age.
The degree to which different social classes live in proximity or not is also an important factor. Even if no adolescent or parent has any wish to arrange marriages by social class, if they live in socially segregated towns then there will be a tendency towards socially homogamous marriages, because meeting chances are higher for individuals living in the same area of a town; hence spatial social segregation will encourage marriages between members of the same social class. The same applies to the organization of associational life. Over the past two centuries in the Western world the number of clubs and organizations has probably grown, and if, say, a soccer club or a youth movement attracts young men and women mainly of the same social class, then that too will create a drive towards socially homogamous marriages. On the other hand, if associational life is socially mixed, that will stimulate social mobility at marriage.
The penultimate factor in the list of determinants of marriage patterns by social class discussed here is the labour market, and in particular the rise of large firms. Bureaucracies such as banks, post offices and the civil service, as well as factories, provide not only work but also spouses. Marriage may occur between people who have met at work, and in large firms young men and women are presented with an opportunity to meet each other, so the degree to which that leads to socially mixed marriages depends on how socially mixed the workplaces are. If nearly all of those at work are from the same social class, say in a factory, then that will lead to social homogamy. In the reverse case it will lead to social heterogamy.
The final determinant refers to the degree of religious or ethnic diversity in a community, the notion being that if a community is split along both social and religious lines, and the religious cleavages are deep, then there will be a drive towards socially mixed marriages. The reasoning here is that the number of marriage partners from the same social class, that is deemed suitable for marriage, is reduced, because some will have a religious background deemed unsuitable. In such a case most people might give priority to marrying someone of the same faith over marrying someone from the same social class.
CONCLUSION
This introduction to what, according to the literature, may be seen as the driving forces behind regional and temporal variations in social mobility patterns has not only been short, and as a result highly stylized, but – based as it is on the existing literature – it also has its lacunae and might give the impression that all forms of social mobility everywhere and at all times have had the same determinants. That would be a false impression, or at least one that is only partially true.
The extent to which trends in the past two centuries, or differences between regions, can be explained by the determinants that have been discussed is an empirical question. We do not yet know what factors prevailed in some societies and what in other societies, nor do we know how much there will remain in each society that we cannot explain and for which we might need either to find new generalized explanations or to bow our heads before the historical record and acknowledge idiosyncrasies.
It would be a false impression, or at least an overstated claim, in another sense too. The literature on intergenerational social mobility does not quite suggest an identical array of driving influences as does the literature on social homogamy or career mobility does. And there is no reason why it should, either. Following in one's father's footsteps or not; embarking on new labour trajectories or receiving a gold watch at the end of a lifetime of service to the same employer; or finding a marriage partner from a similar or perhaps a radically different social origin are not the same phenomena. They can all be said to reflect aspects of social openness, but they do so in different ways and are in part shaped by different driving forces. This survey of the literature has, however, identified similarities, and it remains in large measure another empirical question to what extent the same driving forces have operated for all three forms of mobility and to what extent the determinants were different. One way or another, however, institutional contexts, so it is claimed, will prove to have been of crucial importance in shaping social mobility patterns in history.
ACKNOWLEDGEMENTS
This article has been written under the Advanced Investigator Grant of the European Research Council, ‘Towards Open Societies? Trends, Variations and Driving Forces of Intergenerational Social Mobility in Europe over the Past Three Centuries’, ERCAdG, nr230279 (2009–2014). I would like to thank Ineke Maas and my other colleagues in the Department of Sociology at Utrecht University and at the International Institute of Social History for their comments on earlier versions.
