A cohort is a group of people born at about the same time. Thus, one can speak of the birth cohort of 1951–1960 to refer to people born in that ten-year period. A cohort can also be defined in terms of other significant demographic events. A marriage cohort is a group of people married around the same time; a schooling cohort is a group of people who experienced some educational milestone such as starting the first grade or graduating from high school or college at the same time (for example, the "Class of 2000”); a migration cohort is a group of people who entered or left a country at about the same time; and so forth. However, the most common cohort considered in demographic, social, and economic analysis is the birth cohort (see Whelpton 1954; Ryder 1968; Shryock, Siegel, and Associates 1976, pp. 550–3).

In technical analyses conducted by demographers, birth cohorts are generally defined in terms of some standardized period of time, such as a decade between censuses, but in the social sciences generally it is more common to define birth cohorts in terms of events that were important enough to have shaped the lives of a generation. Thus, the "Depression-era generation” refers to those born in the years from about 1910 through about 1925, who came of age during the Great Depression of the 1930s and fought in World War II. The baby boom generation, or "boomers,” includes those born in the ten to fifteen years of historically high fertility beginning in 1946. Generation "X” – sometimes called the "baby busters” – are those born during the period of unusually low fertility beginning about 1965 and extending through 1980.¹

It is common to categorize people according to their birth cohort and to identify differences in the attitudes and behavior of different cohorts – "generation gaps.”  This is because unique events such as war, famine, economic depression, political struggles, legislative change, and even the advent of television seem to have a lifelong impact on those coming of age at the time they occur. Moreover, behavior is often influenced by what others are doing at the same time. For example, in the 1950s, employed mothers with young children often had to defend their decision to work to family and friends. In the 1990s, it was the "stay-at-home moms” who more often felt on the defensive. The combination of changed circumstances and the social tendency to adjust one's behavior to coordinate with that of one's neighbors means that in our history we sometimes observe radical differences in the behavior of people born just a few years apart from one another. These behavioral differences can show up in a variety of areas including health, educational attainment, marriage and divorce, fertility, employment, and worldview. In some cases, such cohort differences in behavior and attitude characterize a lifetime. A number of recent books attempt to highlight generational differences in attitudes and values. For a detailed depiction of the generation that came of age during the Great Depression and World War II, see Tom Brokaw's The Greatest Generation. William Strauss and Neil Howe propose a larger historical generational cycle, with four generations to a cycle, each of which displays a distinctive world outlook. Meredith Bagby offers a sketch of "Generation X. Thomas Schelling discusses the implications of the fact that in many situations individuals' behavior depends on the behavior of those around them (Schelling 1978; Strauss and Howe 1990, 1993, 1997; Bagby 1998; Brokaw 1998; Howe and Strauss 2000).

This chapter presents data arranged so as to illustrate some of the major differences in the life experience of distinct birth cohorts in American history. As such, it is a reworking of data presented elsewhere. There is great value, however, in presenting data in cohort format, because cohort differences are often invisible in conventionally displayed statistics. Discussion in the other chapters of Historical Statistics of the United States is conducted in terms of what demographers call "period analysis.”  Period analysis focuses on some aggregate of the behavior of people of all ages at one particular point in time – for example, daily newspaper circulation per household. One reason for the predominance of period analysis in Historical Statistics is that most economic and social statistics are collected and presented as a characterization of the entire population or economy at some point in time. For many issues, period analysis is the appropriate research strategy. For example, the most salient characteristic of the Great Depression of the 1930s was the massive failure of the macroeconomy. The Great Depression lasted a full decade and at its depth nearly a fourth of the labor force was unemployed (see Chapter Cb, on business fluctuations and cycles). It had a sizable and permanent effect on people of all ages and all birth cohorts, and it is appropriate to focus on the many differences between this period and the adjacent periods of American history.

For other issues, however, period analysis is inappropriate. Take the example of newspaper readership. Ever since Alexis de Tocqueville wrote in the early nineteenth century, social observers have commented on the connection between newspaper readership and civic engagement. Recently, Robert Putnam has argued that newspaper readership remains a mark of "substantial civic engagement.”  For this reason, it is important to understand the marked decline in newspaper readership that began about 1950. Newspaper circulation per household was 1.24 per day in 1950 and has declined steadily since then, reaching 0.55 in 1998 (series Ae29 and Dg268). Was the decline the result of some force that affected all age groups, much like the Great Depression, or was it something else? Putnam argues as follows:

Newspaper reading is a lasting habit established early in adult life. If we start young, we generally continue. Virtually none of the precipitous decline in newspaper circulation over the last half century can be traced to declining readership by individuals. Virtually all of the decline is due to the by now familiar pattern of general succession…. [T]hree out of every four Americans born in the first third of the twentieth century continue to read a daily newspaper as the century closes just as that generation did decades ago. Fewer than half of their boomer children are carrying on the tradition, however, a fraction that has dwindled to one in four among their X'er grandchildren. Since more recent cohorts show no sign of becoming newspaper readers as they age, circulation continues to plunge as the generation of readers is replaced by the generation of nonreaders. (Putnam 2000, pp. 218–9)

Overall, then, the downward trend in the period data results from successively lower newspaper-readership rates among consecutive birth cohorts. This means that the public policy implications are more serious than they would be if the decline in readership affected all age groups. As Putnam concludes, "reversing that slump will not be easy, since each year the ground is slipping away beneath our feet.”

In principle, one could observe differences in newspaper readership across birth cohorts by studying "cross-sectional” data with a disaggregation by age. Cross-sectional means that data for a single year are disaggregated in a way that allows the user to observe differences in the behavior of people of different ages at one point in time. These cross-sectional data for many different years are another type of data found in many chapters in Historical Statistics of the United States. Like period data, cross-sectional data are useful for many purposes. Putnam's observation, described in the previous paragraph, is one example.

Despite their usefulness, however, period and even cross-sectional data can obscure changes in behavior across cohorts. To illustrate this point, consider Figures Af-A and Af-B, both of which are derived from the same underlying data. Figure Af-A displays cross-sectional data on the proportion of women in the labor force, by age, at seven different census years beginning in 1920. It is tempting to view one of these cross-sections as if it represented the experience of a typical individual as she aged. In this case, a "synthetic cohort” approach would seem to imply one constancy and two dramatic changes in the overall level and life-cycle pattern of women's labor force participation over the twentieth century. The constancy is the rising pattern of women's labor force participation with increases in participation from the teens to the midtwenties and then a decline into the retirement years evident in every census year. One change is an increase in the level of employment at almost every age over the course of the twentieth century. The other is the disappearance of a dip in female labor force participation during the key child-rearing years in a woman's midtwenties through her midforties. In 1920 and 1940, the cross-sectional data seem to imply that women entered the labor force after completion of their schooling, withdrew when they married and had children, and then reentered after their children were in school or had left home. By 1990, the "withdrawal” is virtually eliminated. It seems as if those women participating in the labor force in their midtwenties stick with it for a good thirty years, but begin retiring at a relatively early age (50 years old).

However, as Claudia Goldin (1990) has shown, an entirely different view of change in female labor force participation across the life cycle emerges when the data are rearranged in a way that highlights the actual experience of different cohorts as they aged, as in Figure Af-B. For example, the labor force experience of the cohort born in 1931–1940 is assembled by considering the participation rates of 20- to 29-year-olds in 1960, 30- to 39-year-olds in 1970, and so forth.

In Figure Af-B it is clear that no group of women born in the twentieth century displayed a consistent decline in participation before the age of fifty-five. The apparent withdrawal of women's participation in the labor force during the peak child-bearing years that is so striking in the synthetic cohort does not represent the actual experience of a true cohort. Although the cohorts of 1901 through 1920 exhibit a decline in their middle years, they show the highest participation rates of their lives in their fifties, after their children were grown. For the cohort of 1921–1930, the labor force participation withdrawal during the child-rearing years disappears completely. For the cohorts of 1931–1940 and younger, there is a clear and pronounced pattern of increasing participation at every age, up to the age of fifty-five.

The moral of this story is that in a rapidly changing economy, the synthetic cohort approach based on cross-sectional data can be misleading. Synthetic cohorts give an accurate picture of life course only if cohorts do not differ radically from one another in their behavior. Where cohort effects are strong – that is, where the behavior of one cohort differs substantially from that of earlier and later ones – they can confound true life-cycle patterns. In the case of women's labor force participation rates, marked increases in participation at each age across successive cohorts raised the rates of young workers relative to older workers so much as to make it appear in the cross section that participation rates were falling at older ages. Cohort data make it clear that the opposite is the case.

This chapter presents a small number of data tables arranged to draw attention to differences in the experiences of successive cohorts as they age and also to show what cohort data look like. These tables focus on three different areas of life experience: labor force participation, education, and marital status. The essay also points to shifts in the behavior and experience of different cohorts that can be detected in data displayed in other chapters of Historical Statistics of the United States. These thumbnail sketches are meant as illustrations of the enormous range of intercohort variation in life experience. A full descriptive effort would require the development of many new data series and, therefore, is beyond the scope of this project. Fortunately, the technique for translating cross-sectional data disaggregated by age into cohort data can be applied to many different series.² Even where cross-sectional data by age are not available, one can often guess the cohort patterns by remembering that changed circumstances often have their greatest impact on people who are young adults at the time. For reasons that will become clear, young adults are the ones most likely to embrace new attitudes and behavior.

Back To Top

In a dynamic economy such as that of the United States, the demand for labor is constantly shifting across occupations and industries. Technological and organizational changes mean new products (the automobile in the 1920s), new methods of production (the tractor in the 1920s and 1930s), and shifts in the demand for inputs (more college-educated labor in the 1980s and 1990s). Increase in the sheer size of the economy makes it profitable for some to make a full-time occupation out of an activity that in an earlier era was carried out as a sideline. Real estate services are a recent example. Wars, natural disasters, and mineral discoveries are other developments that can alter labor demand in ways that are difficult or impossible to reverse. One example is the destruction of the American merchant marine during the Civil War (1860–1865). The fleet was owned by Northerners; the Confederate Navy conducted successful raids (Table Eh59–94). As a consequence, shipping business moved to foreign fleets. The American fleet never recovered (see Table Df606–611 and Df736–741). Changing tastes are also a factor. Americans' growing preference for athletic shoes and "casual Fridays” in the latter part of the twentieth century reduced the demand for shoe shines. There are many other examples (see, for example, the essay and tables on occupations in Chapter Ba).

Occupational shifts tend to have strong cohort effects because, as in the case of newspaper readership, occupational choices tend to be made at young adult ages and to remain relatively fixed for life. There is a good reason for this. Most occupations require costly and specialized training before a novice becomes proficient. Even where this training takes place on the job and involves no out-of-pocket expense, it often takes several years of experience to achieve full proficiency. For these reasons, older workers who have already achieved proficiency in an industry or occupation tend to stay with it, even when prospects worsen. Even if relative returns in the industry itself may have fallen, the first industry to which they are attached may offer superior rewards given the costly learning needed to excel at some other occupation. By contrast, new labor market entrants have every reason to select occupations and industries that pay high wages and where future prospects seem good. Thus, shifts in occupational and industrial demand tend to create disjunctures in the experiences of different cohorts.

The decline from cohort to cohort in the share of the workforce engaged in farming is one example. At the beginning of the eighteenth century, agriculture accounted for almost the entirety of the American workforce; at the end of the twentieth century the share was about 2 percent (Table Da-C and Table Ba1033–1046.) This decline in the agricultural workforce presents one of the great paradoxes in the history of the American economy. Whereas the industrial side of economic life is widely acknowledged to be dynamic and increasingly productive, the quantitative record suggests that productivity growth in agriculture was even more impressive. It is ironic that agriculture's relative success in this regard was precisely the reason it shrank as a share of the total economy. While farm productivity and farm output were rising, farm prices were falling. Farm incomes could not keep pace with those in other sectors. Farmers' children discovered more lucrative careers in industry, services, and the professions.

The cohort-to-cohort change in the occupations of women is another dimension of long-term change that is large enough to be visible even in the aggregate data. In 1870, more than 40 percent of both black and white women workers were employed as domestic servants. Among white women, improvements in education – in particular the high school movement – and the expansion of white-collar clerical jobs meant that by 1920 fewer than 10 percent of white working women were employed in domestic service whereas almost one quarter (23.4 percent) worked as clerks and typists. When educational opportunities for black women expanded following the Great Migration, black women also moved into clerical work and their domestic service work declined proportionately (see Table Ba1103–1116 and Table Ba1117–1130).

To observe other changes that may have had similar, dramatic impacts on the life experience of adjacent cohorts, one needs to take a closer look at the detailed occupational data presented in Tables Ba652-813 and Tables Ba1159-4206. It is an exercise that can be both revealing and enjoyable. For example, most people would not be surprised to observe the rise through 1900 and the subsequent decline in the number of blacksmiths. However, not many people are aware that there were more than one million miners in the 1920 workforce, or that in 1900 more than 2.5 percent of the workforce was engaged in the hand production of clothing and hats. Today, the share is about one tenth of 1 percent. No one will be surprised at the meteoric growth in the number of economists over the last half of the twentieth century, but they may be startled at the rapid increase in occupations associated with security, law enforcement, legal matters, religious observance, insurance, and real estate. For most workers in these occupations, the occupational shifts at the national level represent a generational shift as well.

Back To Top

Immigration can affect the size and demographic structure of cohorts. Because the number of immigrants and their characteristics have fluctuated substantially over the course of U.S. history, immigration has had an uneven impact across the different cohorts (see Chapter Ad, on international migration).

People tend to migrate as young adults. Thus, one way to display the relative importance of immigrants across cohorts is to show the share of the foreign-born among those in their twenties. Data in Table Aa185–286 and Aa2026–2077 suggest that immigration had a substantial impact on cohorts born in the mid- to late nineteenth century. For these cohorts, the foreign-born account for about one sixth of the total when these cohorts are in their twenties. The passage of the Quota Acts in the 1920s and then the Great Depression of the 1930s brought immigration virtually to a halt. Indeed, during some years of the Great Depression, more American residents emigrated from the United States than foreigners immigrated to the country. This dramatic cessation of immigration is reflected in an equally dramatic reduction in the impact of immigration on the cohorts born during the early years of the twentieth century. At the nadir, the foreign-born accounted for only about 2 percent of these cohorts when they were in their twenties. The post–World War II baby boom cohort was also only slightly affected by migration from abroad, but the situation is quite different for the small birth cohorts that followed. For the cohort born about 1965, the foreign-born account for approximately one tenth of the total. This resumption of immigration after years of relatively small flows partially offsets the small number of births in the "baby bust” generation born after 1965.

Back To Top

For nearly two and a half centuries, black slave labor was an integral part of the economic, social, and political life of the American South. The most profound consequences of the abolition of slavery were naturally felt by the former slaves themselves – the freedmen. On the eve of the Civil War, in 1860, there were more than four million slaves in the United States. Almost 90 percent of the black population and more than one third of the total population of the Southern states was enslaved (Table Aa2093–2140 and Table Aa2141–2188). Slaves accounted for almost half of the taxable wealth in the Southern states (Table Eh50–58).

The Fourteenth Amendment to the Constitution, passed in 1865, made these four million people suddenly free. For the first time in their lives, freedmen could retain their labor earnings and dispose of them as they pleased. Following Emancipation, freedmen reduced the labor of young children and women (Table Ba11–24 and Ba50–63). Children were enrolled in school instead of being sent to the fields (Table Bc438–446). Fertility fell (Table Ab52–117) and more young children were cared for by their own mothers and fathers (Table Ae128–190). The improvements in quality of life were substantial.³

At the same time, opportunities for the black population remained far more limited than those for whites. Roger L. Ransom and Richard Sutch (1977) blame this failure on the character of the economic and social institutions constructed in the years immediately following the Civil War. Further progress required either a change in these institutions or the departure of blacks from the South. The Great Migration, the movement of blacks from the rural South into the urban North from 1910 to 1950, helped to accomplish both. In the forty years following Emancipation, only slightly more than a half million blacks left the South for the North. In the forty years beginning in 1910, more than 3.5 million did so. This figure refers to net migration, that is, the excess of out-migration over return migration.

In 1900, fewer than 5 percent of blacks born in the South lived in other regions of the country. By 1950, the figure was more than 20 percent (Table Ac1–42) (Eldridge and Thomas 1964; Collins 1997). Because these migrants were overwhelmingly young adults, the impact on the lives of the cohorts who came of age in those years was far greater than these numbers suggest.

The Great Migration expanded opportunities for black Americans in two ways. First, by getting out of the relatively impoverished rural South, blacks gained access to better-paying jobs for themselves and to better education for their children. Virtually all of the decline in the relative importance of farm and farm laborer occupations for black men and women displayed in Table Ba1089–1102 and Table Ba1117–1130 is due to the movement of blacks out of the rural South and into the great cities of the North, Midwest, and West. Much of the improvement in the education of black children over the same years is due to their parents' move away from the South (Margo 1990).

Another consequence of the Great Migration is that it forced Southern whites to improve the quality of the schools and jobs they offered to blacks. With the onset of the migratory flow, blacks suddenly had a new option in life. If conditions at home were unattractive, they could move to the North. Robert A. Margo (1990) documents the substantial improvement in the quality of schools for blacks relative to those for whites in the South following the initiation of the Great Migration.

Back To Top

The size and quality of life of a cohort are affected by its experience of mortality. Across the span of American history, cohorts have had different experiences in this regard. Evidence on mortality experience by cohort is presented in Table Ab656–703.

Two points stand out. One is the huge gulf between life expectancy at birth and life expectancy at age five years in the nineteenth and early twentieth centuries, and the much more rapid improvements in life expectancy at birth relative to improvements at other ages over time. In the nineteenth century, the United States had exceptionally high child mortality levels – approximately one in five children died before reaching the age of five years. Samuel Preston and Michael Haines (1991) show that children of the well-to-do together with those of the poor suffered this plight. Improvements resulted from advances in the medical understanding of infectious disease and from better public health efforts such as sewage systems and water filtration plants to control disease spread. The second point is that year-to-year fluctuations in mortality are much greater for earlier cohorts. Before the development of detailed knowledge about communicable diseases and the ability to control their spread, there were periodic epidemics that killed large numbers of people (see Chapter Ab, on vital statistics). People did not know, from year to year, what risks they might face. Historians have argued that this uncertainty led to a mindset characterized by fear and fatalism. The control of disease nourished confidence and optimism. The transition was accomplished in less than fifty years – and was experienced within the lifetime of cohorts born around the turn of the twentieth century.

Back To Top

Because the United States maintains a relatively small peacetime army and goes to war only infrequently, it has had relatively few veterans in the population relative to the experience of many other countries. Nonetheless the major wars, especially World War II, involved a large share of draft-age males. These circumstances mean that military experience and veterans' status differ dramatically from one cohort to another. One measure of these differences can be obtained from series Aa228–229 and Ed234–235, which can be used to calculate veterans 30–39 years of age as a share of all males of that age group according to their year of birth. This measure records only those who survived their military combat and for this reason understates the impact of wartime experience on affected generations (for combat deaths see Table Ed82–119). Nonetheless, the cohort-to-cohort differences are stark. About half of the men born around 1840 experienced military service (in the Civil War); among the cohorts born between about 1850 and 1880, very few did. The experience of their younger brothers and sons was quite different. Survivors of the U.S. involvement in World War I from the cohorts born in the 1890s represented about one third of this population. Cohorts born in the early twentieth century were most involved. Virtually all men born in the years immediately following the conclusion of World War I had some military experience. Although there was a decline from this high peak, a full generation later, among the cohorts born in the early years of the baby boom, fully half of all men served in the military. Then it ended. Among the cohorts born between 1960 and 1980, military experience was the exception. Fewer than 10 percent of men had been involved by the time they reached their thirties.

Back To Top

The term "baby boom” refers to the temporary reversal in the long-term decline of the crude birth rate following the conclusion of World War II. The baby boom is clearly visible in series Ab40, which displays data for the crude birth rate for two centuries beginning from 1800. The crude birth rate dropped well below the long-term trend during the difficult years of the Great Depression and then rose during the 1950s to levels that had not been experienced since the early 1920s. It is the combination of an unusually small number of births during the 1930s, an unusually large number in the 1950s, followed by a small number again in the 1970s that gives the baby boom its name.

The large baby boom cohort is easy to spot in Table Aa185–286, which displays population by age at each of the decennial censuses beginning in 1850. The substantial long-term decline in fertility (and mortality) reduced the relative importance of successive cohorts at young ages (0–9 years) from almost 30 percent of the total in 1850 to only 16 percent by 1940. As a result of the baby boom, the size of this young age group rose to almost 22 percent by 1960. The continuing numerical importance of the baby boom cohort through the twentieth century is evident in the disproportionate size of the 10- to 19-year-old age group in 1970, the 20- to 29-year-old age group in 1980, and the 30- to 39-year-old age group in 1990. Population projections indicate that the baby boom generation will be a distinct feature of the U.S. population structure through the first half of the twenty-first century.

The direct and indirect consequences of the unusually large baby boom generation were first and most famously analyzed by Richard Easterlin (1962, 1968, 1980). Many are readily apparent in data series that appear throughout Historical Statistics of the United States. For example, the statistics on elementary and secondary school enrollments show pronounced expansions and then contractions as the baby boom generation moved through these institutions (Table Bc7–18). Even college and university enrollments reveal the impact of the baby boom cohort, although the effect is muted because of the rapid rise in college enrollment rates among the generations that preceded and followed the baby boomers (Table Bc523–536).

Employment measures also reflect the labor force entry of the baby boom generation and its maturation over time (Table Ba470–477 and Table Ba478–486). Young male and female workers were especially numerous in the late 1960s and the 1970s. In the 1980s and 1990s, middle-age workers predominated in the labor force. In the early decades of the twenty-first century it will be older workers, both at retirement age and beyond, who will form the largest share. It is especially remarkable that the sheer size of the baby boom cohort predominates even in the case of female workers. As noted earlier, successive cohorts of women have exhibited higher labor force participation rates at each and every age. Others things being equal, such a trend in participation would have meant that successively younger cohorts would outnumber their older sisters. This is not what happened.

Easterlin analyzed the negative consequences of its large size for the baby boom generation itself. To the extent to which young and older workers are poor labor market substitutes, membership in a large cohort will mean more intense labor market competition, lower wages, and higher unemployment than would otherwise be the case. From the point of view of the economy, however, having a large fraction of the total population in the working age groups is a good thing because these are the people who support the young and the old. Development economists measure the importance of the population in the working age groups relative to the total population with the "dependency ratio.”  The dependency ratio is calculated by dividing the sum of people 0–14 years of age and those 65 years of age and older by the number of people 15–64 years of age. The dependency ratio for the United States can be calculated from the population-by-age data from the federal Censuses of 1850 through 1990 (Table Aa185–286) and from the annual population-by-age data for the period 1900 through 1998 (Table Aa125–144). Such calculations reveal a dependency ratio that drops from 0.79 in 1850 to 0.48 by 1940, rises to a local maximum of 0.68 in 1961, after which it falls, reaching a nadir of 0.50 in 1986. Thus, the increase in per capita income between the early 1960s and the late 1980s was in part due to reductions in the dependency ratio. With the aging of the baby boom generation, the dependency ratio will rise again. Improvements in worker productivity will be required to forestall reductions in per capita income.

Back To Top

Although there is a small number of explicit cohort tables in Historical Statistics of the United States, the imaginative and industrious user can apply the approach described in this essay to many of the more standard data series and identify cohort differences across a broad spectrum of American life. The examples presented here were chosen in part because they illustrate a variety of techniques of analysis. They are also great stories. There are other great stories in the data, including some that still await their storyteller.

Back To Top

Table Ba404–416.

For display purposes the means of the age ranges are used and 80 is the value assigned to the category "75 and older.”

Back To Top

Series Af8–14.

For display purposes the means of the age ranges are used and 75 is the value assigned to the category "70 and older.”

Back To Top

Bagby, Meredith. 1998. Rational Exuberance: The Influence of Generation X on the New American Economy. Dutton.

Brokaw, Tom. 1998. The Greatest Generation. Random House.

Carter, Susan B., and Richard Sutch. 1996. "Myth of the Industrial Scrap Heap: A Revisionist View of Turn-of-the-Century American Retirement.” Journal of Economic History 56 (1): 5–37.

Collins, William J. 1997. "When the Tide Turned: Immigration and the Delay of the Great Black Migration.” Journal of Economic History 57 (3): 607–32.

Easterlin, Richard A. 1962. "The American Baby Boom in Historical Perspective.”  Occasional Paper 79, National Bureau of Economic Research.

Easterlin, Richard A. 1968. Population, Labor Force, and Long Swings in American Economic Growth. National Bureau of Economic Research.

Easterlin, Richard A. 1980. Birth and Fortune: The Impact of Numbers on Personal Welfare. Basic Books.

Eldridge, Hope T., and Dorothy Swaine Thomas. 1964. Population Redistribution and Economic Growth, United States, 1870–1950, volume 3, Demographic Analyses and Interrelations. American Philosophical Society.

Goldin, Claudia. 1990. Understanding the Gender Gap: An Economic History of American Women. Oxford University Press.

Howe, Neil, and William Strauss. 2000. Millennials Rising: The Next Great Generation. Vintage Books.

Margo, Robert A. 1990. Race and Schooling in the South, 1880–1950: An Economic History. University of Chicago Press.

Preston, Samuel H., and Michael R. Haines. 1991. Fatal Years: Child Mortality in Late Nineteenth-Century America. Princeton University Press.

Putnam, Robert D. 2000. Bowling Alone: The Collapse and Revival of American Community. Simon & Schuster.

Ransom, Roger L., and Richard Sutch. 1977. One Kind of Freedom: The Economic Consequences of Emancipation. Cambridge University Press.

Ryder, Norman B. 1968. "Cohort Analysis.”  In David Sills, editor. The International Encyclopedia of the Social Sciences, volume 2. Macmillan.

Schelling, Thomas C. 1978. Micromotives and Macrobehavior. Norton.

Shryock, Harry S., Jacob S. Siegel, and Associates. 1976. The Methods and Materials of Demography. Condensed edition by Edward G. Stockwell. Academic Press.

Strauss, William, and Neil Howe. 1990. Generations: The History of America's Future, 1584 to 2069. Morrow.

Strauss, William, and Neil Howe. 1993. Thirteenth Gen: Abort, Retry, Fail, Ignore? Vintage Books.

Strauss, William, and Neil Howe. 1997. The Fourth Turning: An American Prophecy. Broadway Books.

Sutch, Richard. 1975. "The Breeding of Slaves for Sale and the Westward Expansion of Slavery, 1850–1860.”  In Stanley L. Engerman and Eugene D. Genovese, editors. Race and Slavery in the Western Hemisphere: Quantitative Studies. Princeton University Press.

Whelpton, P. K. 1954. Cohort Fertility: Native White Women in the United States. Princeton University Press.

Back To Top