Home Articles Reader Opinion Editorial Book Reviews Discussion Writers Guide About TCRecord
transparent 13

Historical Development of Age-Stratification in Schooling

by David L. Angus, Jeffrey E. Mirel & Maris A. Vinovskis - 1988

Two aspects of age segregation in American educational history are traced--segregation of children from adults due to expansion of public education in the 19th and early 20th centuries, and increasing stratification of children by age within schools due to the practice of age-grading. (Source: ERIC)

Among social historians today, life-course analysis is exceedingly popular because it provides a flexible framework for investigating the lives of individuals. Emphasis is placed on understanding the historical background of cohorts as well as the institutional settings in which they lived. Unfortunately, while the role of norms and social organizations in affecting the expectations and experiences of people is stressed, it is seldom investigated. Instead, the existence of these regulatory norms and social organizations is simply assumed without analyzing their development or their changing impact on the life course of individuals.

A common theme in studies of the life course is the age stratification of American society. Roles and activities within our society are frequently segregated by age. Schools, for example, are designed for the young while retirement and the social security system are intended for the elderly. A few writers acknowledge that American society has not always been age-graded, but little effort has been made to analyze the origins of age stratification and its impact on the life course.

To see how age-grading became so important in our lives, we will examine the development of age-graded schooling. Schools were one of the first social organizations to stress age-grading and they continue to play an important role in shaping the life course today. First we will analyze the establishment of formal schooling in colonial and nineteenth-century America and how this contributed to the demarcation of childhood as a special part of the life course. Then we will investigate the increasing age-grading within schools in the nineteenth and twentieth centuries. By studying the emergence of youth as a period set aside mainly for formal education and analyzing the increasingly close identification of chronological age with grade in school, we will see how age-graded norms and institutions became an integral part of the life course in modern society.


An unstated assumption today is that children are different from adults. In our age-graded society, we usually make even finer distinctions among children on the basis of age. In the past children were not as sharply distinguished from adults. Aries claims that medieval European society regarded children as “miniature” adults. He argues that childhood did not emerge as a recognized separate phase of the life course in Europe until the seventeenth and eighteenth centuries.1

Influenced by the work of Aries, several family historians maintain that colonial Americans also did not differentiate between children and adults.2 Most scholars, however, reject this interpretation, since the Puritans made this distinction.3 The Puritans did believe that young children were more capable of intellectual activity and moral responsibility than we do today. As a result, while childhood was perceived as a distinct and separate phase of the life course, Puritans did not perceive a large intellectual or emotional gulf between older children and adults. There was no sharp or clearly defined boundary, for example, between being a youth and becoming an adult. Although early Americans also acknowledged some differences between young and older children, the life course of children was not highly differentiated. Early Americans paid relatively little attention to the actual age of children and focused more on individual intellectual and physical differences.

New England Puritans believed that everyone should be able to read the Bible. While a few local schools were established in the mid-seventeenth century, most of them were intended only for the minority of advanced male students planning to attend Harvard University. The primary responsibility for teaching children the alphabet and how to read rested with parents, who sometimes preferred to send their children to a nearby dame school.4 The dispersed nature of settlement in the Chesapeake area as well as the relative lack of interest in education meant that even fewer schools were established in the South.5 Most children in early colonial America therefore were taught the rudiments of reading by their parents or a neighbor rather than by a teacher in a formal public or private school.

The number of public and private schools increased greatly after the American Revolution. The Puritans had always stressed the importance of literacy from a religious perspective. The creation of the new republic necessitated an educated male electorate, which then also stimulated educating women as the mothers of the next generation of voters.6 Concerns about the social tensions caused by nineteenth-century urbanization and industrialization as well as the necessity of bringing civilization to the West also spurred educational expansion and reform.7 While regional differences in education persisted, the major expansion occurred in the Midwest and South as those areas initially lagged behind the Northeast in school facilities and attendance.8 Throughout most of the nineteenth century gender differences in education were small, but blacks consistently received less schooling than whites.9

The rapid expansion of schools in the nineteenth century reflected in part the increased unwillingness of parents to educate their children at home. Although adult male and female literacy had greatly increased, parents preferred more than ever to send their children to schools—especially if the cost of that education could be spread among all taxpayers. Some educational reformers advocated the expansion of schools to compensate for the inadequate child rearing provided by lower-class and immigrant parents, whom they regarded as incapable of providing—or unwilling to provide—a suitable education at home.10

The net result of all of these changes is that by the mid-nineteenth century, most children in the United States received at least some training in a school. In rural areas children of all ages frequently attended a one-room schoolhouse while in urban areas schools were usually subdivided into several classes. Going to school in the country was a seasonal activity contingent on the need for agricultural labor at home; in urban areas schools remained open for most of the year. Consequently, there were significant differences in the amount of education available to children in rural and urban areas.11

At first, schools accepted children of all ages. Since colonial Americans regarded young children as capable of intellectual training at an early age, it was not unusual to see three- or four-year-olds in the same classroom as teenagers. The infant school movement in the 1820s emphasized the importance of early education and encouraged sending young children to schools. An estimated 40 percent of three-year-olds were in Massachusetts schools in 1840.12

A reaction against early childhood education occurred in the mid-1830s. Physicians like Amariah Brigham began to argue that premature intellectual stimulation of young children weakened their growing minds and eventually led to insanity. Concern over an alleged decline in the quality of American family life produced a burgeoning literature on child rearing and the responsibilities of parents, particularly mothers. Schoolteachers and administrators joined the crusade against early education, since it was difficult to educate toddlers along with the older children. As a result, schools gradually enacted regulations prohibiting young children from attending and thereby set a minimum age at which children could enter.13 The introduction of kindergartens in the second half of the nineteenth century encouraged sending some younger children to schools, but most parents and educators thought that children should not be in school until age seven or eight.

At the same time that a minimum age for school entry was being established, educators were concerned about the presence of older boys in the country schools. Horace Mann estimated that nearly 10 percent of all rural Massachusetts schools in 1840 were terminated prematurely because the older teenagers literally threw the teacher out or wrecked the schoolhouse. Therefore, school administrators worked to lengthen the school year and increase the regularity of attendance so that children could complete their primary education earlier. With the creation of public high schools in the second third of the nineteenth century, the length of schooling was extended. Since students often entered high school at age twelve or thirteen and most did not stay long enough to graduate, nineteenth-century children typically completed their formal education by their mid-teens.14

Thus, the nineteenth century witnessed the expansion of schooling and the compression of childhood education into a shorter age span. The length of the school year was extended and emphasis was placed on getting everyone to attend school regularly between the ages of seven and thirteen. As a result, the life course of children of these ages was increasingly defined by their school attendance and educational experiences.

During the past century, schooling in the United States has become an even more important factor in defining the life course. The proportion of the school-age population in schools has increased dramatically. In 1870 approximately one-half of the population ages five to nineteen were in school; today about nine out of ten children in that age category attend school. The large differential between black and white school attendance rates has all but disappeared.15

Schooling is now the typical experience of even younger children and of young adults. Today, more than one-third of those ages three to four are in school and almost one-fourth of those ages twenty to twenty-four are in school.16 The median number of school years completed by adults ages 25-29 has risen from 10.3 years in 1940 to 12.9 years in 1985.17

The average length of the public elementary and secondary school year has also increased during the past one hundred years. Though school terms in cities grew shorter, from nearly year-round school in the 1830s to a forty-week year by the turn of the century, rural terms expanded from as little as three months to as many as nine; thus, the average school year in the United States expanded from 132 days in 1870 to 179 days in 1970. Combined with an increase in the regularity of attendance, the average number of days attended per enrolled public school student rose from 78 in 1870 to 162 in 1970.18

As schooling became such a dominant part of the experience of American youth, full-time work lost its earlier importance. In colonial America young teenagers either worked on farms or were apprenticed to artisans and tradesmen. Child labor was normal and expected. As schools replaced home education in rural areas, older children continued to work during the summer, attending school only in the winter. In urban areas, where schools remained open most of the year, children usually either attended school or entered the labor force. Only a few children combined schooling with full-time work outside the home.

As the emphasis on education increased, child labor laws were enacted to remove young teenagers from the workplace and to put them into the classroom. While many of these laws date from the 1870s, most observers suggest that the means for their enforcement were not in place until the 1890s. The average minimum age for leaving school increased from fourteen years five months in 1900 to sixteen years in 1930 in those states that had enacted a compulsory schooling law. The combination of more restrictions on child labor and the increase in school attendance led to a decrease in teenage labor force participation. In 1900 about 35 percent of females and 70 percent of males ages fourteen to nineteen were working while the comparable figures for 1950 were 25 percent and 40 percent.19

Thus, the life course of children today is much more defined by the experience of schooling than ever before. Indeed, in the United States school attendance is nearly universal for children between the ages of five and seventeen and is being expanded for both younger and older ones as well. In addition, the school year has become longer and there is even talk about increasing it further to make it more comparable to other industrialized countries such as Japan. While the proportion of teenagers full-time in the labor force has been greatly reduced in the twentieth century, the number of older students with part-time jobs has risen so that the sharp division between schooling and work has been somewhat diminished.


If throughout the nineteenth and into the twentieth century children were increasingly segregated from the activities of adults and assigned to that special institutionalization of childhood, the school, what role did age play in locating them within the school? The typical school at the opening of the nineteenth century enrolled pupils of widely varying age, some as young as three, some as old as twenty, all taught in the same room by a single teacher. Today, children enter kindergarten at five years of age, and with few exceptions, pass through the elementary and middle school grades, one year for each grade. Between the ninth and twelfth grades they either drop out of school to enter the work force after reaching the minimum legal age, or remain to graduate, again progressing one grade for each year. Thus, the internal organization of the school is fairly strictly age-graded. Children and teens spend the years from five to seventeen or eighteen within age cohorts of an extremely narrow range.20 How, when, and why did this remarkable, yet seldom remarked upon, shift occur?


In the one-room district schools of rural, antebellum America, all children of “legal school age” were eligible to attend.21 Legal definitions of school age were features of state school laws coming into existence in the 1820s and 1830s and were most often functional for the purpose of determining the amount of state financial support each district was eligible to receive annually.22 This definition varied widely, from Iowa’s five to twenty to Massachusetts’ five to fifteen. On the low end of the scale, children sometimes went to school at three or four, but as we suggested earlier, this practice declined after 1830. At the upper end of the scale, most children stopped going to school before they reached the legal maximum age.

Some modest age segregation was introduced by the common practice of dividing the school year into two terms, a summer school, attended largely by children from four to five to about ten or twelve, and a winter school, attended by those from ten or twelve to twenty. The summer school was typically taught by a local girl in her teens, emphasizing a few rudimentary subjects (spelling, reading, ciphering), and providing much time for play and organized games. Its primary purpose seems to have been keeping the little ones out from under the feet of adults during the busy season while inculcating some school “readiness” skills. The winter school was more often taught by an “outsider,” a man or older woman who could handle the disciplining of teenage boys and offer instruction of a more advanced nature in such subjects as history and geography, as well as the “basics.” Exceptions to this age grouping were common, however, since many young children who lived near the school attended the winter session as well as the summer.

Within the school, instruction was organized according to what was called the “recitation” method. The school day was divided into many short periods, each devoted to a particular subject and a level of study within that subject, such as beginning spelling, intermediate reading, advanced arithmetic, and so forth. During each of these brief periods, and while the majority of students sat at their “desks” to study, the several children who were performing at the appropriate subject and level would group around the teacher’s desk to “recite,” reading passages, answering questions, or working problems on command. Many illustrations of these “old district schools” suggest that age (or at least size) had little or nothing to do with the makeup of these recitation groups. Since children differed widely with respect to when they had begun school, how many terms they had attended, how regular their attendance had been, how much time they could find at home to study, and so forth, not to mention differences in talent and motivation, age was a poor predictor of which child would be studying at which level in each subject. Teenage boys who had grown up out of the reach of a school might find themselves learning the alphabet alongside children of four or five.

Even in “extracurricular” activities, the reminiscences and diaries of country schoolteachers suggest only a modest amount of age segregation. The ubiquitous ball games, tag games, and hiding games were played by all but the very youngest children. The older boys had some “activities” of their own, sometimes as forms of rebellion against the teacher, but also sometimes in concert or even conspiracy with the teacher.23


The situation with respect to age segregation within schools was at first little different in the cities and towns. Of course the much greater density of population meant that a school district would contain many more children than a single room could accommodate or a single teacher could handle. But the multiroom, multiteacher schools of the antebellum period were only roughly graded. In Detroit, for example, when the public school system was created by statute in 1842, two grades of schools were set up, primary schools, enrolling boys live to ten and girls five to twelve, and middle schools, enrolling children over those ages. Vestiges of rural influence remained. Primary teachers were female, middle teachers were male and were paid more, the recitation method of instruction predominated, and the rooms within a school building were called “schools” rather than classes or grades.

This began to change quite rapidly after 1850. In 1847, John D. Philbrick, principal of the Quincy Grammar School of Boston, worked out a system of “graded instruction” that went well beyond the rough division into primary, middle, or grammar schools.24 As it developed and spread over the next couple of decades, the system had two features. First, the subjects to be taught were standardized with the material arranged in an orderly sequence of increasing difficulty. A group of subjects arranged in this way was then called a “course of studies.” Second, teaching was subject to a “division of labor” in which each teacher would be responsible for teaching particular segments of material from the hierarchical arrangement of subjects. Once a school building was organized in this way, it was called a “graded school,” and the children were assigned to different teachers depending on their level of past accomplishment.

In the most advanced graded schools, the whole course of studies was divided into segments representing one year’s work on the part of the average child. There might be eight or nine such segments. The ideal was that a child would enter school at five or six, progress through the required work at the rate of one segment, or grade, per year, completing the course at fourteen or fifteen, passing into the high school by examination or leaving school at that time. Passage into each level was based on attaining a particular score on an examination or maintaining a particular average on all subjects throughout the year. In actual fact, this orderly progress seldom occurred.

It is important to emphasize that it was the curriculum and the work of the teacher that was graded, not the children. They were “classified,” that is, placed into classes based on what they had learned. Age was irrelevant. In fact, public school leaders in Detroit, describing their new graded system to parents and taxpayers, explained, “On the Union plan several hundred scholars in different departments are placed under the general superintendence of a principal teacher, and subject to his classification. This classification is based upon the degree of attainment in studies, regardless of age or condition in life.”25 It is wrong then to assume, as a number of writers have, that the graded school was either intentionally or actually organized according to chronological age. As we shall see, there was only a loose relationship between children’s ages and their grade placement in the urban graded school of the nineteenth century.

The graded school was immensely popular. As one commentator later put it, “By 1860 the schools of most of the cities and large towns were graded. By 1870 the pendulum had swung from no system to nothing but system.“26 There were a number of advantages to this form of school organization. In grouping children according to their level of accomplishment, the teacher could work with larger groups instead of with two or three at a time, as in the district school. Eventually, a system of “class” instruction emerged, much like the “chalk and talk” system that dominates today. Under this instructional arrangement, discipline became easier to maintain, and because of grade-level specialization, teachers could be effective with less training. This in turn allowed city school boards to hire more female teachers, which they could do at little more than half the salary of a male teacher. With the children seated at fixed desks in neat rows, the number of children per classroom could be quite large, and school buildings were built to contain eight or ten rooms, instead of the four rooms more common in the 1830s.

These features of schools would hardly be thought of as “advantages” today, and indeed they have been heavily criticized as elements of the process of “bureaucratization” that are said to have made urban schools such sterile and oppressive places for children.27 When understood against the background of the demographic conditions of the time, however, they were highly advantageous—even essential. As Angus has observed elsewhere, the most important urban educational problem of the nineteenth century was “overcrowded schoo1s.“28 Caught between exploding numbers of school-age children on the one hand and miserly common councils on the other, urban school boards were desperate to simply provide a place in school for every eligible child. Any organizational scheme that could stretch scarce dollars, and at the same time give some semblance of order and progression, was seen as a blessing.


Criticism of the graded school began to be heard as early as 1872. W. T. Harris, superintendent of schools in St. Louis, in his annual reports for 1872, 1873, and 1874 and in an 1874 speech to the National Education Association, attacked two features of the system.29 He argued that annual promotions held back students who might be moving through the curriculum at a more rapid pace, and that many students, failing to be promoted a second time, withdrew and were permanently lost to the school. He proposed a more flexible grading scheme with each “grade” divided into several levels under the same teacher and with the brighter pupils promoted into the next level every ten weeks. The plan produced considerable debate among superintendents, and though it was not generally adopted, it started a trend toward half-yearly or quarterly promotions rather than the more traditional practice of annual promotions. More importantly, school leaders all over the country began to think about and work on alternative forms of instructional organization.30

One was William J. Shearer, who as superintendent in New Castle, Pennsylvania, and Elizabeth, New Jersey, perfected a plan of frequent promotions based on subject mastery, flexible enough to allow each student to be at different grade levels in different subjects. Throughout the 1890s, the plan was touted in the press and adopted in a few other cities. In 1899, Shearer described his plan in a book entitled The Grading of Schools, including a Full Explanation of a Rational Plan of Grading, “the first book written upon this important phase of school organization. ” Shearer defined the graded school as one in which “a definite course of study” is taught and in which “the pupils are roughly classified according to their supposed ability to do the work of a given year: and each class is placed in charge of a teacher, who is expected to give the same lesson to all members.” While he accepted the idea that the graded school was superior to the ungraded school of earlier times, it had numerous defects, chief among which was the fact that it did “not properly provide for the individual differences of the pupils.“31

No one dares deny that the children of every grade differ widely in age, in acquirements, in aptitude, in physical endurance, in power of attention, in home advantages, in the rate of mental development, in the time of entering school, in regularity of attendance, and in many other ways affecting their progress. . . . Because of the manner of grading and promoting, the graded school of today keeps all the children of each grade in intellectual lock step, not only month after month, but year after year, for their whole school lives. . . . Is not individuality of more importance than evenness of grading?32

Apparently not, for the idea that was to dominate the attention of Progressive educators in the next two decades was not the individuality of the child, but the inefficiency of school systems in which thousands of children failed each year, some time after time, leaving the schools in their early teens having completed only a few elementary grades. Rather than educational lockstep, it was these two features of urban schools—called “retardation” and “elimination”—that would become the focus of reform efforts.


Early elimination from school, what today we call the dropout problem, had received some attention in the nineteenth century.33 The problem of retardation, defined as the number of children “overage” for their grade, and its relation to elimination was first brought to attention in the 1904 report of the superintendent of schools in New York City, Dr. William H. Maxwell. Maxwell gathered and reported data on the ages of the pupils in each grade of the New York schools, finding that 39 percent of all elementary pupils were “retarded” (that is, overage) for their grade. The finding caused a stir among big city superintendents, and several of them, notably those of Philadelphia, Chicago, St. Louis, and Detroit, followed Maxwell’s lead in collecting and commenting on similar data. A number of academics also became interested in the question. Edward Thorndike began collecting age/grade data from as many cities as possible; the Russell Sage Foundation agreed to fund a “backward child” study of urban schools; George Strayer designed a set of forms to carry out a nation-wide survey; and Lightner Witmer, the founder of “America’s first psychological clinic” and of the journal The Psychological Clinic, began to devote space to articles on the problem.34

The school-efficiency movement based on age/grade statistics began in earnest with the publication of Thorndike’s The Elimination of Pupils from School in 1908. Analyzing the statistics he had gathered with great care, Thorndike concluded in part,

At least 25 out of 100 children of the white population of our country who enter school stay only long enough to learn to read simple English, write such words as they commonly use, and perform the four operations for integers without serious errors. A fifth of the children (white) entering city schools stay only to the fifth grade. . . . Of the children entering the public schools of our more favored cities over half probably never have a man teacher. Less than 1 in 10 graduate from the high school. Only about a third graduate from an elementary school of seven grades or more. . . . One main cause of elimination is incapacity for and lack of interest in the sort of intellectual work demanded by present courses of study.35

When this study appeared, it was greeted by widespread newspaper comment, numerous editorials chiding schools for the “shocking degree of inefficiency” disclosed by the data, and a storm of defensive criticism on the part of superintendents who denied the accuracy of Thorndike’s findings.

An even bigger stir was created by the publication a year later of Leonard P. Ayres’s Laggards in Our Schools: A Study of Retardation and Elimination in City School Systems. Ayres had been general superintendent of schools for Puerto Rico and chief of the division of statistics of the Insular Department of Education before taking the job as chief investigator in the Russell Sage “backward children investigation.” The book not only stamped Ayres as the chief apostle of the school-efficiency movement, it raised the ante on the level of statistical sophistication a well-informed educator would need to follow the new developments.

Ayres reported his findings under three headings: the condition, the causes, and the remedies. With respect to the amount of retardation and elimination in the schools, Ayres found that “the general tendency of American cities is to carry all of their children through the fifth grade, to take one half of them to the eighth grade and one in ten through the high school,” though there was enormous variability between cities. Since the number of overage children in the grades exceeded those who were under age by a wide margin, Ayres concluded that “our courses of study as at present constituted are fitted not to the slow child or to the average child but to the unusually bright child.” An array of “causes” was analyzed—late entrance, irregular attendance, illness, physical defects, the nationality of the child, the sex of the child—but none was shown to be a “preponderant” cause. As remedies, Ayres called for better compulsory attendance laws, better medical inspection, courses of study better fitted to the average child, more flexible grading, and above all, better collection of school statistics.36

Ayres’s book created an even greater storm than Thorndike’s. Over the next decade, lucky was the superintendent who was not challenged by the local press to report the numbers of children who were overage for grade and to begin to show progress on this important new measure of “school efficiency.“37 The age/grade table, displaying the numbers of children of each age within each grade, rapidly became a routine aspect of the annual report of many districts.38 Calculations were made of the “money cost of the repeater,” though this concept was hotly debated.39 A study of a random selection of one hundred city school reports from 1905 to 1912 showed that three-quarters of them made some mention of the concept of retardation and that half presented “adequate” statistics on the problem.40

Raymond Callahan has developed an interpretation of the school-efficiency movement that gives primacy to such factors as the “vulnerability” of city administrators to the pressures of the business community.41 Callahan’s account downplays such other factors as the search for legitimacy and confidence on the part of the newly emerging, university-based field of educational psychology and the new developments in educational measurement and statistics. The importance of these can be illustrated, first, with the case of Lightner Witmer and, second, with some of the debates that occurred between those who were within the efficiency movement, such as Thorndike and Ayres.

In 1910, in an article in his own journal, The Psychological Clinic, Witmer described his interest and involvement in the age/grade question. After tracing the origin of interest in the education of the retarded to Itard’s work with the wild boy of Aveyron, through the work of Itard’s pupil Edward Seguin, to the development of special institutions for children of various degrees of retardation in the United States, Witmer argued that

Seguin’s message to the teacher of normal children has been neglected and “retardation” has come to be regarded as synonymous with an incurable mental defect. Feeblemindedness, imbecility, and idiocy are but classes or grades of retardation. In these severe forms of retardation the mental defect usually rests upon some incurable brain defect. In milder forms of retardation the child may be backward because of no organic defect but because of functional nervous disease, of inadequate or improper nutrition, of defects of sight or hearing, or perhaps merely because he has not been sent regularly to school or may not have been subjected to satisfactory home discipline. Many of these cases of retardation are curable under appropriate physical and educational treatment, that is to say, the child can be restored to the mental and physical status of the normal child of his age.42

To Witmer, the age/grade statistic held out the best chance to measure the extent of “backwardness,” or “pedagogical retardation,” in the general school-age population and to explore its causes and treatment. To this end, he encouraged superintendents in Philadelphia and Camden to collect age/grade data even before they knew of Maxwell’s effort, and devoted considerable space in his journal to the debate between Thorndike, Ayres, Roland Falkner, Oliver Cornman, and others.

When Thorndike gathered data from city school reports for his 1908 elimination study, the best available statistics were tables of enrollment by grade and the numbers of children in the schools at each whole year of age. To answer such questions as the degree of overageness, the rate of progress through the grades, the typical age and grade of elimination, and so forth, he was forced to make certain estimates based on assumptions, particularly concerning the number of children entering a system in a given year. Ayres’s study was based on better age/grade data for New York City, but he had no better data than Thorndike had for other cities. Nevertheless, Thorndike’s assumptions, and thus his estimate of the extent of retardation, were sharply attacked. Over the next few years, the argument was joined by others, on Thorndike’s side by his graduate students and on Ayres’s by his colleague in Puerto Rico, Roland Falkner.43 As better statistics became available, the arguments shifted. Should educators be concerned with overageness for grade or only with failure to make normal progress regardless of age? What are the prime causes of retardation and/or elimination? What is an adequate system of “child accounting”?44 What is the best age for a child to begin school?

If the newly emerging profession of psychology saw in the concept of agegrading a measurement tool that could lend scientific legitimacy to their work, the same can also be said of the emergent field of scientific school administration. Administrators in the Progressive era were enthralled by statistics and numbers. During these years, many fervently believed that the “truths” contained in tables and graphs could transform education from a jumble of well-worn cliches and practices into a full-fledged science grounded in objectivity and research.45 Few numbers proved to be more important than those detailed in age/grade tables. Arthur Moehlman, a leading advocate of this scientific administration and “child accounting,” described the utility of age/grade tables:

The gross determination of curricular fit is made through semester or annual promotions and failure reports, age-grade placement, and agegrade progress studies. The grades organization is based on the assumption of orderly progress, one grade per year, through the school. When the theoretical rate of progress is exceeded, the result is acceleration, and where the theoretical rate is not attained the result is retardation. The relationship of children to progress, regardless of other factors, is secured from a study of failure to make grades in terms of the total possibility. A study of conditions within a building or a total school district may be made from a chart showing a two-way distribution of age in relation to grade. The age-grade study shows the condition and indicates the point of attack. [These charts provide] a quick picture of conditions prevailing at any time.46

But if the age/grade table provided a “quick picture” of a school’s quality, what standard of age/grade “fit” should schools attempt to attain and how should it be measured? Here, too, numbers formed the language of the assessment of educational quality. In the early work, both Thorndike and Ayers had used a two-year age range for each grade as “normal.” For example, children who were six or seven and in the first grade were considered at grade level. This was made necessary, in part, because school districts collected their statistics at different times of the year. As child accounting became more standardized, administrators realized the value of an early fall data collection, since this greatly reduced the percentage of overage students, and some districts shifted to a one-year or one-and-a-half year standard. This shift in the standard makes cross-time comparisons more difficult. Some educational leaders, such as Ellwood P. Cubberley of Stanford University, declared that a normal rate of progress would result in about 70 percent of all students being at grade level with about 15 percent being accelerated and 15 percent being retarded.47 This is only about 10 percent better than New York City reported in 1904, but at that time, the two-year interval standard was being used, and the ages of the students were collected in June.

However measured, these proved to be tough standards. Applied to George Strayer’s national survey of 1911, few cities and towns would have met them. Over the next decade, as cities continued to publish age/grade tables and strive for higher efficiency, few were able to show progress, even using a fairly broad definition of the proper age-for-grade. New York, for example, which had led the nation in identifying and attacking the problem of overaged pupils with its report of 39 percent overage in 1904, still reported an overage rate of 31 percent in 1922, Superintendent Ettinger referring to “the promiscuous distribution of children of all ages or of children of a given age through the grades.“48

As such results became publicized, it began to appear that school leaders were unable to meet the very standards they had set. Adding to administrators’ problems was the fact that business leaders, key allies in the campaign for efficient school systems, frequently complained about the obvious inefficiency of schools that forced students to take eight or nine years to complete six years of elementary education.49 As David Tyack notes, “In an age that worshiped efficiency, over-aged students and school leavers were signs of malfunctions that required analysis and correction.“50

By the early 1920s, educators were employing a variety of means to extract themselves from this age-grading “trap.” For example, some readily shifted the blame for skewed age/grade distributions onto the students themselves. They claimed that the enormous increase in urban school enrollments was responsible for the high rates of failure and overageness, particularly the large numbers of foreign-born or rural children whose previous education was "inadequate.“51 A subtle means for actually improving age/grade distributions was for school administrators to change the way that systems determined whether a student was at grade level. As one critic later noted, “In most school systems a high rate of non-promotion was considered an admission of educational inefficiency. Therefore, school authorities tended to decrease the number of non-promotions by their accounting practices.“52 School systems, for example, began passing large numbers of children with “conditional" promotions. Administrators recorded these students as promoted and at grade level, even if the students were later demoted. Administrators also changed the listing of students who were demoted but who subsequently left school from “nonpromoted” to “dropped.” Finally, new students who were placed below grade level were listed as “transfers” rather than overage.53

While these changes in child accounting did improve the age/grade distributions of many school systems, other changes that school reformers introduced during this era proved to have an even greater impact on “normalizing” age groupings. These changes included the expanded use of ungraded classes, special education programs, IQ testing, tracking, ability grouping, and even substantial curriculum reform designed to make success in school easier to achieve for larger numbers of students. As William Ettinger, superintendent of schools in New York City, asserted, “I believe that the rapid advance in the technique of measuring mental ability and accomplishments means that we stand on the threshold of a new era in which we will increasingly group our pupils on the basis of both intelligence and accomplishment quotients and of necessity, provide differentiated curricula, varied modes of instruction, and flexible promotion schemes to meet the crying needs of our children.“54 The age/grade movement was a catalyst for the widespread adoption of all these educational practices.


We can best see the development of this process by looking closely at one school system. In 1907, the Detroit public school system conducted its first age-grade survey. The number of students at grade level in 1907, in 1914, and again in 1917 was nowhere near Cubberley’s norm of 70 percent. For example, 49 percent of sixth graders in 1917 were overage for grade.55 Such figures forced the system toward a major reassessment of its educational program.56 School leaders moved to correct the problem, ordering studies on the causes of retardation and overageness, creating a department of special education to absorb handicapped students, and introducing “a differentiated course of study to meet the different abilities within a grade.“57 By 1922, Detroit officials were estimating that about 3 percent of the system’s students should be in special education or ungraded classes.58

A central feature of this effort was mental testing and ability grouping. In September 1920, Detroit began testing all entering first graders and developed an elaborate system of programs and tracks to meet the “needs” of the various ability groups.59 In addition to these efforts, school leaders sought to reorganize both the structure and the content of elementary education. As one Detroit official put it, “The appalling situation with respect to retardation and elimination in many of the larger centers of population had been pointed out by scholars some years before. The plain inference was established that a discipline which so great a percentage of pupils were unable to master was not a situable discipline for the majority of children.“60 Thus, in the early 1920s, Detroit adopted a version of the Gary Plan for its elementary program. Known as platoon schools, this program combined the teaching of traditional elementary subjects such as reading and arithmetic with enrichment courses such as music and art while also providing classes in physical education, dramatics, and so forth.61 These changes were, in part, intended to improve the progress of students through the grades. Besides restructuring elementary education, Detroit’s school leaders also transformed secondary education, shifting to a 6-3-3 pattern of organization for the system and adding vocational and practical courses in junior high schools (called intermediate schools) as a means of reducing overage students in grades seven and eight.62

By 1922, in a spirit of self-congratulation for these efforts, Detroit boasted that 74 percent of its elementary students were either at or above grade level.63 This may have represented considerable progress for Detroit, though it is not clear whether this calculation was based on the one-year or two-year interval.


While school systems such as Detroit’s may have led the way in achieving high rates of at-grade-level students, the nationwide trend toward that goal proved to be quite uneven. Schools in states outside the industrial Northeast and Midwest usually lagged behind in implementing age-grading and in reaching “normal” age/grade distributions. During the 1930s, many of these states still reported very high rates of failure and consequently high levels of overageness among elementary pupils. In West Virginia, for example, school officials found in the 1934-1935 school year that the failure rate in some counties ran as high as 75 percent. In 1938 Maryland teachers failed 22.4 percent of white first-grade boys and 14.4 percent of white first-grade girls. Figures were even higher for black children. That same year in Tennessee about 16 percent of the state’s black children and 12 percent of its white children were required to repeat a grade.64 Rural schools, even in the Northeast and Midwest, were also slow in reaching “normal” rates of students at grade level. A 1935 study of students in 292 one-room rural schools in Michigan, for example, found an average of 59 percent of the students at grade level and an average of 19 percent overage, with the percent of overage students rising in every grade.65

Even in major urban school systems, the rates of nonpromotion and the numbers of overage students remained high into the 1930s. In New York City, for example, retardation rates for elementary children within the fifty-four subdistricts of the school system ranged from a low of 16.2 percent to a high of 40.8 percent in the 1934-1935 school year. School officials in Chicago reported nonpromotion rates for first graders in 1936 running between 10 and 20 percent. In 1939, in St. Louis 30 percent of the city’s first graders were not promoted, an increase of 11 percent from 1930.66

Despite the persistence of large numbers of overage students, and notwithstanding the difficulties of cross-time comparison, there does appear to have been a significant movement toward higher rates of students at grade level during the 1930s. One study of the Virginia public schools, for example, analyzed state level data from 1924 to 1937 and found that “the median [age] and distribution of ages in the individual grades have dropped—a phenomena [sic] which suggests improved grading and promotion practices. Presumably, there is less retardation of older children, ‘and children in the same grade are more nearly of the same age.“67 Similarly, a study of changes in retardation rates in seven major American cities showed a definite decline in the percentage of nonpromotions and, we can assume, a higher percentage of at-grade-level students between the early 1920s and 1938-1939.68

It is important to take note of the fact that, throughout the three decades in which the age/grade movement became a powerful catalyst for a broad range of Progressive reforms, educators never abandoned the notion that children’s grade placement should have something to do with what they had learned. While they were prepared to experiment with a host of devices to improve age/grade “fit,” and thus the efficiency of their systems, they did not take the simple expedient of placing children in grades based on their chronological age alone. This step remained to be taken by still more “advanced” Progressive thinkers.


The movement to get the vast majority of students at grade level received a final, and ultimately triumphant, push in the 1940s. These years witnessed a shift in conventional educational wisdom as important as the one associated with the introduction of age-grading itself. Much as age-grading changed the definition of a quality school system from one with high rates of failure to one with high rates of promotion, so in the 1940s, educators began to adopt the idea that automatic promotion, or as it would later be called “social promotion,” of virtually all students was the sign of true educational quality.


As early as 1931, a committee of school administrators in New York City urged that “grade standards be abolished and that pupils in the elementary and junior high schools be promoted on the basis of school attendance and chronological age.“69 The essence of the administrators’ plan was to “virtually do away with non-promotions and put the city school system on a 100 percent promotion schedule.“70 While New York did not adopt the plan at that time, by the late 1930s the concept of automatic promotion was gaining legitimacy in educational circles. In 1936, for example, A. C. Krey predicted that “universal school enrollment up to the age of sixteen, if not also beyond this point possibly even to the age of twenty, will continue and that this will be accompanied with an administrative policy approaching automatic promotion.”71

As in the earlier campaigns to make more frequent promotions (1870s to 1890s) and to adjust age-grading norms to account for individual differences (192Os), advocates of this policy focused their attacks on traditional, inflexible promotion practices. Willard Elsbree, a leading advocate of flexible promotions, railed that “in most communities, it is still taken for granted that each grade has established achievement norms; that the curriculum should be geared to these standards; and that pupils should master in orderly sequence the subject matter and skills prescribed in the elementary program. Failing the mastery of subject matter and skills, the pupil should repeat the year’s work."72 In place of these practices, educators sought a “plan of education that emphasizes the development of social human beings rather than the performance of certain routines which may over the course of years happen to give some factual information which will be retained in after life and may incidentally develop the characters of children so they will be able to meet the world.“73

In many ways, the call for automatic promotion was the logical culmination of a number of elements of Progressive educational thought, which stressed the need to (1) accommodate individual differences within the classroom, (2) educate the whole child, particularly with regard to the psychological health of children, and (3) educate all American youth. Relying on what they claimed was the best educational research, social promotion advocates argued, first, that failing children was a clear indication that a school was not motivating children properly. If children failed the fault lay not in the children but in the outmoded organization and practices of the school. Furthermore, students who failed a grade generally did not improve in achievement. Failure, and the threat of failure, simply did not prove to be a good motivator.

A second line of argument was that students who failed were psychologically damaged by the experience and that this psychological damage had profound, negative educational consequences. Studies showed that children who failed had greater self-doubts (what later researchers would call lower selfesteem) and a strong sense of inferiority and disgrace. Once branded a failure, these students evidenced increased feelings of antagonism and bitterness toward the schools and teachers, often leading to discipline problems in the classroom.

Finally, students who failed frequently became high school dropouts, coming to see an early exit from school as the best solution to their ongoing educational and behavioral problems. Advocates of social promotion thus saw the policy of failing students as another manifestation of the undemocratic character of traditional educational practices that “forced” students to abandon their education entirely, denying them “equal educational opportunity.“74

Many educators, especially those in leading urban systems, eagerly accepted these ideas. As early as 1938, a National Education Association survey of 23 cities found that most superintendents favored some form of social promotion.75 By the early 1940s such major school systems as Minneapolis and New York City put social promotion policies into place.76 In a 1955 study, Roger Lennon and Blythe Mitchell compiled data from five national reports that contained information on age-grading. They concluded that “policy with respect to promotion of elementary school pupils has undergone considerable modification in the past 30 years or so, moving unmistakably in the direction of so-called ‘social’ or chronological-age-based promotion and away from promotion premised on attainment of any fixed level of achievement.“77 The introduction, development, and consequences of social promotion in urban schools can be illustrated by the experiences of Philadelphia and New York City.


Philadelphia, which had played a key role in the earliest debates on age-grading and adopted a modified form of social promotion earlier than most major districts, provides a good example of the development of the practice. In the late 1930s, Philadelphia administrators introduced a policy of “continuous pupil progress.” This policy set in motion a major reorganization of elementary education in which

emphasis has been placed on group work and different ways of classifying pupils. Courses of study and time allotments are viewed as guides. . . . Diagnostic and remedial teaching is resulting in development of all kinds of self-teaching, self-directing, and self-checking arrangements. Individualization of instruction is recognized as the basis of a program of continuous pupil progress. Informality in classroom procedure is becoming more prevalent. The entire administrative setup is more flexible.78

The most important aspect of this program was that “grade, for all practical purposes, has become synonymous with school age.“79

The policy, however, generated intense opposition. By the late 1940s, Philadelphia school officials were forced to defend social promotion from charges that it had contributed to a loss of standards, rigor, and accountability (on the part of both teachers and students). One group of critics claimed that as a consequence of this policy, “some pupils who should still be in grade school are in high school and many high-school pupils graduate without fundamentals.80 In 1947, the city’s board of superintendents issued a statement denying that the school system had ever adopted a policy of 100 percent promotion. At the same time, however, the board staunchly defended the practice of “continuous pupil progress.” The statement by the board specifically denounced past practices that provided educational opportunities only for students strong enough to survive a system of “rigid, predetermined requirements.” The board argued that the only way to encourage all children to reach their full potential, to truly provide “equality of educational opportunity,” was to drastically decrease the number of failures and allow children to progress at their own rate. Given these goals, the board concluded “that chronological age is the most satisfactory basis for general class grouping.“81


Events in New York City demonstrated that the adoption of automatic, or “continuous,” promotion was not of itself sufficient to create the high degree of age-stratification that characterizes contemporary schools. Between 1942 and 1946, New York replaced a decades-old system of twice-yearly admissions and promotions with a policy of continuous promotion. Students would be admitted only in September, promoted only in June; a clear assumption was that nonpromotion would be extremely rare. In fact, by 1948, the promotion rate for the elementary schools, which had reached a citywide average of 90 percent early in the century, moved to 99 percent.82

Despite this, the administration continued to express a concern about age/grade relationships. Tables and text on age of pupils, age/grade status of pupils, and rate of progress of pupils continued to appear in the annual reports well into the 1960s. Over the years, as a result of the reforms discussed earlier, the percentage of overage students had steadily declined, but the district now faced a new problem. “Since 1923 when attention was concentrated on the problems of overageness and retardation the proportion of overage pupils has been reduced from 29.9 percent to 5.3 percent in 1948. . . .83 While the problem of ‘overageness’ has been brought under control there are indications in the data . . . of a new problem, that of ‘underageness.’ “84 How could this have happened?

As a school district’s rate of pupil promotion approaches 100 percent, the only factor that can affect the age-range of the pupils in each grade is age of entry. In 1948, the New York State school code entitled anyone between the ages of live and twenty-one to attend school. The state’s compulsory attendance law required everyone from seven to sixteen to be enrolled. Theoretically, then, children might enter the first grade as young as five and as old as six years, eleven months, assuming no one violated the law. In actual fact, the policy of the New York City Board was to admit children as young as four to the kindergartens and to admit no one under live years, four months to the first grade. Also, a few parents must have held children out of school past the seven-year limit.85 Given a September data collection and a two-year standard for “normal” (i.e., six or seven for the first grade), it is easy to see why few pupils would be classified as overage and many as underage. The “problem” was further compounded by the postwar baby boom and, possibly, by an increased interest in early child education Between 1942 and 1948, the number of enrolled four-year-olds increased by 163 percent and the number of five-year-olds by 78 percent.86 Reflecting on this state of affairs, the superintendent’s remarks were a powerful echo of the early days of the age/grade movement. “The outstanding characteristic of the data . . . is the variability, as to age, of pupils in the same grade. In the fourth grade, for instance, the ages of pupils range from seven years to fourteen years inclusive. The standard age for fourth grade is nine years. Similarly for each of the other grades.“87

This is not to suggest that all the efforts expended over the years in New York and other places had had no appreciable effect on the degree to which schools were stratified by age. Quite the opposite. There can be no doubt that the distribution of children’s ages in the grades had narrowed, and clustered more strongly around the mean. The key factors in approaching the current level of age-stratification in school were the adoption of social promotion and a decrease in the range of age at entry. Nationally aggregated data suggests that as more and more of the nation’s school districts adopted such policies after 1940, schools became more age-stratified. From 1940 to 1960, there continued to be incremental movement in this direction; however, a comparison of age/grade relationships in the period 1964- 1966 with those in 1983 shows little further change.88


In this study we have reached a number of important conclusions about age stratification and the life course. We have found that age segregation in American educational history has two aspects. First is the simple segregation of children from adults that occurred largely because of the expansion of public schooling in the nineteenth and early twentieth centuries. Second, and for our purposes more significant, is the fact that there has been increasing stratification of children by age within schools due to the practice of age-grading.

The most startling aspect of age stratification in schooling is the relatively recent date of its widespread implementation. As we have shown, it was not until the 1920s and 1930s that leading urban school systems succeeded in getting most children at grade level, and they did this by an assortment of devices that included special classes, ability grouping within classes, and a host of child accounting “tricks.” Rural school systems lagged considerably behind urban ones in approaching normal rates of at-grade-level students. Today’s taken-for-granted level of age/grade homogeneity was not approximated in many schools before the 1940s and probably not in most schools until the 1950s.

Equally important is the way in which age-grading served as a catalyst for a variety of Progressive educational practices including such efficiency-oriented practices as child accounting, intelligence testing, ability grouping, and tracking. Of these practices, social promotion remains one of the most controversial.

After age/grade relationships became a measure of school efficiency in the early twentieth century, educational leaders worked diligently to improve their schools on this measure. Their intention was not to age-stratify the young, but to discover more effective ways of grouping children for maximum learning. With the introduction of the idea of social promotion in the late Progressive period, age-stratification itself became the institutional goal. This shift of intention may be the one aspect of age-stratification that has had the most serious consequences for the nation’s schools.

Whether to hold students back or keep them with their classmates remains a volatile issue. Should the grade placement of children be based, at least in part, on whether they have learned or failed to learn the expected material? Is automatic promotion an example of the schools’ alleged lack of interest in “excellence”? Or does it stem from a deep regard for the psychological well being of the child and does it reflect an abiding commitment to “equal opportunity”? While these issues will divide the educational community for some time to come, it is important to see both how perennial the debate over grouping children for instruction has been and how recent the social promotion “solution” to the question really is.

Cite This Article as: Teachers College Record Volume 90 Number 2, 1988, p. 211-236
https://www.tcrecord.org ID Number: 471, Date Accessed: 1/23/2022 7:05:54 PM

Purchase Reprint Rights for this article or review
Member Center
In Print
This Month's Issue