View sample Social Demography Research Paper. Browse other social sciences research paper examples and check the list of research paper topics for more inspiration. If you need a religion research paper written according to all the academic standards, you can always turn to our experienced writers for help. This is how your paper can get an A! Feel free to contact our research paper writing service for professional assistance. We offer high-quality assignments for reasonable rates.
The description and analysis of inequality and social diﬀerentiation are fundamental components of social demography. In this context, a socially diﬀerentiated population is one characterized by a variety of subgroups that are distinct according to generally persistent social criteria such as race, ethnicity, religion, language, and so on. Inequality, on the other hand, is a hierarchical concept and is used here to refer to diﬀerences in the actual possession of resources of various types or in access to such resources. Resources are in the form of political power or inﬂuence, income, schooling, health care services, etc. Diﬀerentiated groups may also be extremely unequal but the idea of socially diﬀerent need not imply inequality. This research paper brieﬂy reviews several of the reasons why demographers study diﬀerentiation and inequality and then goes on to discuss how four diﬀerent theoretical perspectives have been applied to the problem. The ﬁrst emphasizes cultural explanations; competing somewhat with it are resource-based explanations which focus particularly on diﬀerences in opportunities and constraints. Cutting across both of these are reference group theory and life-course analysis.
Need a Custom-Written Essay or a Research Paper?
Academic Writing, Editing, Proofreading, And Problem Solving Services
A perennial demographic question is the extent to which diverse or unequal groups exhibit marked diﬀerences in demographic behavior. Are there substantial race, ethnic, or religious diﬀerences in marital or fertility behavior, in mortality, or in patterns of migration? One reason it is important to study such diﬀerentials is that they indicate the extent to which the overall demographic performance of a population is a function of subgroup diversity. How big an impact a subgroup’s behavior can have on overall population measures depends on the distinctiveness of that group’s behavior weighted by its relative size. Although size counts for much, if the group’s behavior is unusual enough, it can still have a substantial impact on overall measures. For example, using never married women, aged 30–34, as a rough indicator of marriage delays in the US, 22 percent of all women of that age were never married in 1998. However, the proportion was much higher for African American than white women—47 as opposed to 17 percent. The diﬀerences were so large that even though African American women were only 14 percent of all women in this age group while whites were 80 percent, the proportion never married was pulled up by ﬁve points—or almost 30 percent—for all women as compared to whites (US Bureau of the Census 1998).
Documenting subgroup diﬀerences in demographic behavior also provides important insights into the demographic concomitants of inequality and a socially diﬀerentiated population. It also poses analytical questions about the underlying determinants of these diﬀerences since group membership is usually an indicator of a variety of more proximate causes of an observed demographic pattern—e.g., the role of unequal access to medical care in ethnic group diﬀerences in mortality. A classic example in social demography is the inﬂuence of marriage timing on fertility in early industrializing populations. Given women’s circumscribed reproductive period, delayed marriage in a noncontracepting population typically leads to smaller completed family size, while an earlier marriage promotes higher fertility. Thus, in rural areas in many parts of Western Europe, where inheritance of the farm was a precondition for marriage, marriages were often greatly delayed. On the other hand, factory workers in the cities usually married much earlier as their economic fortunes were tied to the developing wage economy rather than the traditional inheritance system (Anderson 1971).
2. Varying Theoretical Perspectives
The remainder of this research paper will focus on the impact of inequality and social diﬀerentiation on demographic behavior by examining four theoretical approaches to the problem. Together these underlie most research in social demography; hence the discussion will begin by a brief introduction to the general thrust of each perspective before applying them to issues of social diﬀerentiation and inequality. Two distinct and often competing theoretical approaches have been particularly inﬂuential in social demography; these are cultural compared to resource explanations of demographic behavior. Cross-cutting them are two additional perspectives—reference group theory and life-course analysis.
2.1 Socioeconomic Constraints And Opportunities
Resource based explanations focus on how diﬀerences in resources aﬀect demographic behavior—e.g., differences in social or economic constraints and opportunities. Thus, a major feature of the Theory of the Demographic Transition is the argument that economic development raises the cost of children and reduces their economic value to families, thereby promoting the transition from the relatively high fertility characteristic of most preindustrial societies to the typically low fertility of industrialized societies. By focusing on diﬀerences in resources, this approach also lends itself well to analysis of the determinants and consequences of inequality.
In the resource-based approach, diﬀerences or changes in goals or preferences among individuals or groups are frequently not examined in any great detail. In fact, neoclassical economists have typically argued that preferences are ﬁxed across time and do not systematically vary among individuals, thereby permitting changes in demographic behavior to be unambiguously attributed to changes in constraints. However, for most social demographers, such an assumption is generally too strong and variations in preferences are usually considered an important, if often understudied, source of diﬀerences in people’s behavior.
2.2 Cultural Factors
A second approach views ‘culture’ as the most important factor inﬂuencing demographic behavior. The nature of cultural factors being considered are highly variable, ranging from the narrowly to the broadly conceived—from norms regarding the permissibility of divorce or the number of children a couple should raise all the way to more complex and sometimes highly abstract constructs such as religious or basic value systems. For example, it has been argued that both earlier and more recent European fertility transitions are largely the result of a major change in the ‘ideational’ system marked by the growth of secular individualism among other value changes. While some recognition is given to the inﬂuence of economic factors, their importance is often minimized by scholars operating in this tradition. The cultural approach tends to emphasize diﬀerences in goals and values and hence seems more oriented towards explaining socially differentiated behavior rather than behavioral diﬀerences based on an inequality of resources. However, it is often recognized that certain cultural patterns arise out of conditions of inequality and may even contribute to its persistence (Parsons and Goldin 1989).
2.3 Reference Group Models
Cutting across these general perspectives are two additional approaches that have been inﬂuential in explanations of demographic behavior. The ﬁrst is the use of reference group models. The idea here is that people use selected others whose behavior or characteristics provide positive—or sometimes negative— models for themselves. How this approach is implemented partly depends on whether the researcher’s primary orientation is cultural or resource based. If cultural, the emphasis is more likely to be on reference groups as normative models to emulate. If the investigator is more resource oriented, the emphasis is likely to be on relative economic status concerns.
2.4 Life Course Approach
A second inﬂuential cross-cutting perspective has been the lifecycle life course approach. Lifecycle analysis is the pioneering version of this approach and focuses particularly on age-related characteristics and behavior. It can be applied either to an individual—an individual’s career or family cycle, for example—or to a group—the family’s developmental cycle. Because of the scarcity of life history data, lifecycle analysis has often had to restrict itself to age patterns measured at one or more points in time (using cross-sectional data on the behavior of diﬀerent age groups, i.e., ‘synthetic’ cohorts) rather than being able to follow individual life histories over real time. The later developing life course perspective adds a particular emphasis on such life histories, although it too has sometimes had to settle for cross-sectional data. The availability of longitudinal data has increased in recent years, permitting more extensive analyses of life histories. Some life course researchers have also placed great emphasis on normative issues as well, a major one being that there are norms governing the sequencing of life course transitions—for example, that there is a norm against marrying before completing school (Elder 1977, Hogan 1978).
While all these perspectives, singly and in combination, have been generally important in demography, the discussion below will focus on their use in analyses of the role of socioeconomic inequality and diﬀerentiation in demographic behavior. Although often viewed as competing, neither the resource nor cultural perspective alone can do an entirely satisfactory job of explaining patterns or changes in demographic behavior. Both approaches proﬁt by incorporating elements from the other.
3. Cultural Diﬀerences And Demographic Behavior
Probably the most extensive research project which came to espouse a cultural explanation of demographic behavior was the Princeton European Fertility Project which examined the transition from high to low fertility occurring over much of late nineteenth and early twentieth century Europe. By using provincial-level data, the project set out to test several propositions of Demographic Transition Theory, one of which was that fertility declines were a consequence of economic development. A major conclusion of this body of research was that the observed fertility declines appeared to be only weakly related to conventional measures of socioeconomic development, such as the proportions in a province who were literate, employed in agriculture, lived in large cities, etc. In fact, in most European countries the fertility transition appeared to have begun at nearly the same time, despite considerable geographical variability in economic and social conditions. Hence, this too seemed to be a strong indication that the fertility transition was not an adaptive response to the changing conditions brought about by economic development. However, areas which shared similar cultural characteristics, such as language and religion, exhibited diﬀerent rates of fertility change compared to areas at a similar economic level but with diﬀerent cultural characteristics (Coale and Watkins 1986).
To explain these ﬁndings, an innovation and diffusion hypothesis was invoked, involving elements of reference group as well as cultural theories. The argument is that the deliberate control of completed marital fertility was generally unknown in Europe throughout most of its history although large families were not highly desired. However, late in the nineteenth century certain inﬂuential opinion leaders in local areas not only became innovators in the use of contraception but also in promoting the view that family size could and should be legitimately controlled. Their views and practices provided a behavioral model and were, accordingly, rapidly adopted by others. Such changes were more likely to occur within cultural groups, deﬁned, for example, by language and religion, partly because of the ease of within-group communication. In addition, shared norms and values increased the inﬂuence innovators were likely to have within the group, thereby fostering a rapid diﬀusion of innovations. Moreover, some cultural groups were more open to change than others. In Belgium, for example, traditional Catholic doctrines were much more powerful in the Flemish-speaking areas than in the French which were more secularized. Throughout much of Europe, France being a notable exception, fertility changes were more rapid in Protestant than Catholic areas.
The Princeton Project’s contention that the European fertility transition was only weakly related to economic development has been subject to a variety of criticisms. The argument heavily depends on two empirical generalizations—ﬁrst, that the adoption of fertility control was an innovative behavior and, second, that the fertility transition began at roughly the same time over most of Europe despite considerable regional variations in the pace of industrialization. However, because the kind of data collected during this historical period were severely limited, indirect statistical measures had to be developed as indicators of these behavioral patterns. Recent work on the validity of these measures, some by an original contributor to one of the measures (Trussell), indicates that they will not reliably detect the start of the transition until it is well underway and hence arguments about the simultaneity of transitions all over Europe, regardless of the level of economic development, are not reliable. Moreover, the statistical measure of the proportion of couples who were fertility controllers requires a fairly large number of them in order to be detected by the measure, casting doubt on the ability to determine whether fertility control was actually an innovation at the times indicated (Guinnane et al. 1994). In addition, recent analyses of German data, employing more sophisticated multivariate statistical methods than typically used in the Princeton project, have found that indicators of economic development were strongly related to declines in fertility during this period; however, although a cultural factor, such as religion, was important in predicting fertility Levels, it did not have an eﬀect on changes in fertility (Galloway et al. 1998).
While there is increasing doubt that cultural factors alone can explain the European fertility transition, the continuing social policy goal of reducing fertility in currently developing societies has kept alive a strong interest in more rigorously developing an innovation and diﬀusion model. However, recent work in this area has placed a greater emphasis on the value of the process as a mechanism for explaining changing fertility patterns in conjunction with changing socioeconomic constraints rather than instead of them (Montgomery and Casterline 1996). Moreover, in contrast to the provincial-level data the Princeton project was necessarily limited to, detailed microlevel research involving network analysis is being carried out in several less developed countries to determine the particular social interaction and networking mechanisms involved in promoting the diﬀusion of contraceptive use and declines in fertility.
In general, although there is little doubt regarding the importance of investigating the eﬀect of cultural diﬀerences on demographic behavior, in practice it is not always so easy to pin down the causal role of cultural factors. The easiest but probably also the least sophisticated approach is to assume that if there is a simple correlation observed this constitutes reliable evidence of a causal relationship between demographic behavior and membership in socially diﬀerentiated groups such as those deﬁned by religion, ethnicity, social class, metropolitan residence, etc. Sometimes, the existence of cultural determinants is deduced as the residual explanation when other variables, such as socioeconomic ones, do not explain much of the between-group diﬀerences. However, to do so makes the unrealistic assumption that all other relevant variables are included in the analysis and are also properly measured. Moreover, given the preliminary nature of some of the theories under investigation as well as the uneven quality of demographic data and weaknesses in current measurement capabilities, attaching speciﬁc meanings to unexplained variance is always a risky procedure.
Another problem involved in opting too quickly for purely cultural explanations is that cultural identities are usually related to economic characteristics. Hence multivariate analyses are essential for separating out the role of economic constraints from cultural characteristics. Even then, the nature of the causal process is not always clear. For example, education is more highly stressed among Chinese than Latin American immigrant populations in the United States, suggesting a causal sequence between ethnicity, educational attainment, and socioeconomic status rather than one of these factors alone dominating the process. This may indicate a culturally related higher value placed on education among the Chinese—because of a Confucian tradition, for example—but it may also partly reﬂect diﬀerences in the socioeconomic origins of the two immigrant groups—one having a greater representation of professionals while the other coming from a much poorer socioeconomic background. Hence, sorting out the nature of the causal process is often complex and the data available to accomplish this are not always readily available.
The magnitude of the eﬀect of cultural diﬀerentiation on demographic behavior can also be quite diﬀerent depending on whether one is discussing Levels or changes in behavior. For example, religion usually has historically had a consistently sizable impact on fertility in the cross-section but if the religious composition of an area does not change markedly over time, and the eﬀect of any given religious identiﬁcation remains constant, religion’s role in changes in demographic behavior may be minimal. However, if different cultural groups vary in their openness to change introduced from the outside, as is sometimes argued, then cultural diﬀerences could indeed have a substantial impact on change.
4. Socioeconomic Constraints And Opportunities
Social scientists who primarily focus on the role of resources and constraints in demographic behavior have often ignored how institutional and normative commitments may constrain individual choices, both in terms of aﬀecting preferences and also by limiting what are considered the legitimate means for realizing these preferences. In this respect, neoclassical economists have often chastised sociologists for their apparent willingness to posit a plethora of attitudes and norms to explain every observed behavior, thereby trivializing the analytical process. Nevertheless, it is possible to accept the idea of systematic variations and changes in preferences as a reasonable analytical tool and to apply it in a disciplined fashion in conjunction with resource constraints—that is, without positing a diﬀerent or changing norm or value to explain every observed variation in demographic behavior. Hence the view that one must choose either a cultural approach or a resource constraint approach is unnecessarily conﬁning intellectually.
4.1 Relative Economic Status Models
A major example of the integration of the idea of using varying lifestyle aspirations in conjunction with varying resource constraints to explain demographic behavior is found in Richard Easterlin’s work on relative cohort size in the US. Easterlin hypothesized that intergenerational shifts in relative cohort size produce intergenerational inequalities in people’s economic position which, in turn, aﬀect age at marriage, fertility, and a number of other demographic responses. In addition, he maintained that lifestyle aspirations were learned in the parental household so that children raised in aﬄuent households would develop higher-level consumption goals than those raised in poorer households. He combined these two views by arguing that cohorts who were entering young adulthood in the early postwar period were small due to low fertility during the Great Depression and were also raised in less aﬄuent households, resulting in relatively modest consumption aspirations. However, because of this depression cohort’s small size, they were in a very favorable labor market position as adults which encouraged early marriage and childbearing, as well as larger families—in short, they produced the postwar baby boom. The baby boomers, in turn, were raised in these aﬄuent house-holds and acquired high consumption aspirations; however, because of their large size, their labor-market position was not as good as that of their parents and this led to delayed marriage and lower fertility, resulting in the baby bust. In short, the Easterlin hypothesis posited that a feedback eﬀect was produced by this process and created self-generating demographic cycles. Due to intergenerational shifts in relative economic status and consumption aspirations, small cohorts begat large ones which, in turn, produced small cohorts (Easterlin 1987).
Empirically, the status of Easterlin’s theory has been criticized on a number of counts. A major one is that the postwar demographic changes were primarily driven by ‘period’ rather than ‘cohort’ factors. For example, annual birth, marriage, and divorce rates have tended to ﬂuctuate together among larger and smaller cohorts alike, suggesting that there was something about particular historical periods that inﬂuenced behavior rather than cohort size. Nevertheless, the relative economic status aspect of the model remains interesting because the existence of inequality can provide a motivation for adaptive behavioral changes to occur; thus it is an intrinsically dynamic type of model.
Easterlin’s hypothesis focused only on intergenerational inequalities; however, reference group and relative economic status models can also be applied within generations. One example comes from the rapid postwar rise in US married women’s paid employment so that by 1990 around 70 percent of wives, aged 25–54, were in the labor force compared to less then 25 percent in 1950. There is little evidence that this transformation was precipitated by major changes in norms or values. While opinion poles and other attitudinal studies indicate that approval of married women’s employment became increasingly positive with time, attitudinal changes did not precede the changes in employment behavior but rather lagged behind them. A far more inﬂuential factor appears to have been the postwar economic expansion which greatly increased job opportunities for married women because the demand for women workers in female dominated occupations far exceeded the size of the population group that had traditionally been the backbone of the female labor force—young single women (see Oppenheimer 1970).
In addition to substantial changes in labor-market conditions, a type of innovation and diﬀusion model combined with a relative economic status approach might further clarify why the employment shift has taken place so rapidly. The rationale is that the rise in married women’s employment changes how families are stratiﬁed by reducing some previous socioeconomic inequalities and creating others. When married women’s employment outside the home was relatively rare, a family’s socioeconomic position was primarily dependent on the husband’s economic status, although earlier in the twentieth century the work of the family’s adolescent and young adult children was also important at certain periods in the family’s developmental cycle. As wives’ employment became more common in the postwar period, however, many families within the same socioeconomic group, but without working wives, could increasingly be put at a relative economic disadvantage, thereby generating inequalities within groups which had previously been at roughly the same socioeconomic level, based on husbands’ economic position alone. On the other hand, wives’ employment can reduce the relative economic advantage of some families who might previously have been at a somewhat higher socioeconomic level. Hence, wives’ employment in some families but not others creates both an objective and subjective instability in how people experience the stratiﬁcation system. But it also creates a model of how to deal with these instabilities—the entry of other wives into the labor force. Wives already working also provide a source of information on the nature of the job market for others contemplating entering the labor force, as well as providing models of strategies for how wives and mothers could cope with work and family responsibilities.
The impact of this process should, if anything, have increased over time. As long as there were few wives employed, their work would have relatively little impact on both intra and intergroup reference comparisons and the easier it would be to dismiss their behavior as ‘deviant’ or the ‘exception that proves the rule.’ However, as the proportion of wives working rose and became increasingly full-time, then their example would become more salient, thereby increasing its destabilizing eﬀect on reference group comparisons. Moreover, the more wives who worked, the fewer nonworking women there would be around to socialize with, share childcare responsibilities with, etc., hence, the lifestyles as well as the relative socioeconomic position of families with nonworking wives would be increasingly aﬀected. All this should have contributed to the continued rapid expansion of married women’s employment in the postwar era in addition to any changes in labor market conditions (Oppenheimer 1982).
Although not easy to operationalize, due to data limitations, relative economic status models have attractive theoretical features for analyzing change. They posit changes in lifestyle aspirations which aﬀect behavior but which nevertheless remain embedded in a stable preference structure—maintaining one’s socioeconomic position in the community. They also involve feedback eﬀects that foster a continuation of the diﬀusion process whose momentum may even increase rather than peter out over time. Furthermore, the existence of feedback eﬀects means that not all those changing their behavior need experience the same initial motivating factors that the earlier innovators were exposed to; whatever its original cause, the behavior of innovators itself creates an additional set of pressures that inﬂuences others, in turn, to shift their behavior and attitudes as well. Nor need such an amalgam of a relative economic status model with an innovation and diﬀusion model operate independently of social norms. For example, in the American case, although there continued to be a general disapproval of married women’s employment long after it had already substantially increased, even early on in the process there was evidence of a considerable ﬂexibility in attitudes, depending on the circumstances. Thus attitudes were very negative in the 1940s if it was speciﬁed there was only a limited number of jobs but much more permissive if a wife worked in order to facilitate a marriage (Oppenheimer 1970). Hence, the conditionality of many norms permits a certain amount of innovative behavior which may ultimately modify the normative structure more generally. In this case, changing socioeconomic conditions encouraged an increase in wives’ employment which was facilitated by norms viewing working to achieve family goals as a legitimate exception to the general disapproval of married women’s employment. The ultimate result, however, was not only a permanent change in employment behavior but a radical reduction in the general disapproval of wives’ working as well.
4.2 Socioeconomic Inequalities And The Life Course
While cross-sectional analyses are major sources of information on socioeconomic inequality and its demographic concomitants, they often miss an important dimension of inequality: the variation in socioeconomic well-being over the life course and how this may systematically diﬀer among socioeconomic groups as well as change over time. However, there is a substantial body of work—primarily in historical demography—that directly addresses these problems within a life course context.
An inescapable fact of human biology is that individuals’ economic productivity varies over the lifecycle. For example, very young children and the very old have few productive capabilities. The nature of the economic system also inﬂuences these variations and, along with it, the individual’s earnings position. If work involves considerable physical eﬀort or health risks, a reduction in productivity may happen at a relatively young age so that the most economically productive part of the career cycle occurs fairly early in adulthood. On the other hand, if skill is the major job characteristic required, relatively high productivity may last well into old age; however, in this case, lengthy training requirements during an individual’s youth are likely to reduce productivity then. Research on middle-class English, European, and American families and on the working classes around the turn of the nineteenth century not only reveals the substantial general economic inequality between these groups but the considerable diﬀerences in how relative aﬄuence or deprivation was experienced over the life course, along with the resulting diﬀerences in demographic responses. In the case of the middle classes, while earnings might ultimately be substantial, early in the career cycle they were usually inadequate to permit a young man to marry and set up a household in the style considered appropriate to his social station in life. The common pattern was therefore for men to delay marriage until they were economically established and had proved themselves capable of ‘properly’ supporting a wife and family (Banks 1954, Alter 1988).
In contrast to the middle classes, the career-cycle pattern of earnings was far diﬀerent for industrial workers, both in Britain and many other nineteenth century industrializing populations. In addition to their generally much lower earnings, the lifecycle period where earnings were greatest and most assured was in young adulthood rather than middle age; hard physical labor, the risk of debilitating disease or accidental maiming or death in the far more dangerous work settings of that era all raised the risk that the middle and older years would be ones of declining or lost productivity. Several adaptive strategies seemed to have been extremely common in such populations. One was a major reliance on the earnings of children to help maintain or improve the family’s economic stability over its developmental cycle. Another was an early timing to the start of both marriage and childbearing, thus placing these transitions during the period when working class men’s earnings were at their greatest; in addition, this increased the likelihood that at least some children would be old enough to go out to work and contribute to the family’s income as the father aged and entered into a higher risk period. However, the downside to this strategy was a frequent pattern of life-cycle poverty or, at the very least, of relative deprivation—for example, when there were several children in the household but none old enough yet to go out and work: this problem was particularly acute if the earnings of the father were lost or much reduced. A diﬀerent kind of life-cycle poverty typically occurred after all the children had left home and the parents were too old to engage in extensive employment. As a consequence of this life-cycle poverty pattern, the economic situation of individuals at any single point in time was a poor indicator of the frequency with which poverty or relative economic deprivation might be experienced over their entire life course (Anderson 1971, Haines 1979, Rowntree 1922).
While such strategies might well be characterized as rational adaptations to the life course circumstances faced by many working class families, they were also cultural in nature. Historical research indicates that they had a major normative component; children were placed under a strong obligation to work and help improve the family’s economic well-being and they were often allowed to keep very little of their earnings (Parsons and Goldin 1989). Although their work might also improve their own well-being while in the household, such a pattern could nevertheless work against children’s long-term self-interest because it tended to reduce the amount of schooling they received and hence their adult socioeconomic status. Moreover, delaying or even foregoing marriage was not entirely uncommon because of the obligations adult children had to help support the family and/or help care for younger siblings, especially if one or both parents were lost (Hareven 1982).
Work on contemporary populations shows that diﬀerent socioeconomic groups have also been experiencing quite contrasting career development patterns and that this can have signiﬁcant demographic consequences. For example, consistent with their traditionally important and normatively prescribed economic role in the family, it was still true in the 1980s that American men were much less likely to marry while in an unstable labor market position with low earnings. How large a cumulative impact this had on their marriage timing, however, depended on the length of time a young man remained in such an ‘immature’ career state and this varied considerably among racial and schooling groups. Compared to men with one or more years of college, a much higher proportion of high school graduates, as well as dropouts, were experiencing a far more diﬃcult and prolonged transition to a relatively stable career position, leading to substantial educational diﬀerences in marriage timing. African American males were having particular diﬃculties and their marriages were much more delayed than that of whites, as indicated by the never married ﬁgures cited earlier. These ﬁndings also suggest that, given the considerable rise in economic inequality in the US starting in the mid- 1970s, increasing career-entry diﬃculties may have played a signiﬁcant role in the overall rise in the age at marriage, not only for those reaching adulthood during the 1980s but for earlier and later cohorts as well (Oppenheimer et al. 1997; see Family Theory: Competing Perspectives in Social Demography).
- Alter G 1988 Family and the Female Life Course. University of Wisconsin Press, Madison, WI
- Anderson M 1971 Family Structure in Nineteenth Century Lancashire. Cambridge University Press, Cambridge, UK
- Banks J A 1954 Prosperity and Parenthood. Routledge & Paul, London
- Coale A J, Watkins S C (eds.) 1986 The Decline of Fertility in Princeton University Press, Princeton, NJ
- Easterlin R A 1987 Birth and Fortune. University of Chicago Press, Chicago
- Elder G H 1977 Family history and the life course. Journal of Family History 2: 279–304
- Galloway P R, Lee R D, Hammel E A 1998 Urban versus rural: fertility decline in the cities and rural districts of Prussia: 1875 to 1910. European Journal of Population 14: 209–64
- Guinnane T W, Okun B S, Trussell J 1994 What do we know about the timing of fertility transitions in Europe? Demography 31: 1–20
- Haines M 1979 Industrial work and the family life cycle, 1889–1890. Research in Economic History 4: 449–95
- Hareven T K 1982 Family Time and Industrial Time. Cambridge University Press, Cambridge, UK
- Hogan D P 1978 The variable order of events in the life course. American Sociological Review 43: 573–86
- Montgomery M R, Casterline J B 1996 Social learning, social inﬂuence, and new models of fertility. Population and Development Review 22 (Suppl. 5): 151–75
- Oppenheimer V K 1970 The Female Labor Force in the United University of California, Berkeley, CA
- Oppenheimer V K 1982 Work and the Family: A Study in Social Academic Press, New York
- Oppenheimer V K, Kalmijn M, Lim N 1997 Men’s career development and marriage timing during a period of rising Demography 34: 311–30
- Parsons D O, Goldin C 1989 Parental altruism and self-interest: child labor among late nineteenth-century American families. Economic Inquiry 27: 637–59
- Rowntree B S 1922 Poverty: A Study of Town Life, new edn. Longmans, Green, London
- US Bureau of the Census 1998 Marital Status and Living Arrangements: March 1998. Current Population Reports, P20-514.