Growing Chaos in the Lives of Children Youth and Families: How Can We Turn It Around?

Urie Bronfenbrenner
Professor of Human Development and Family Studies and of Psychology
Cornell University

This chapter consists of three parts that lay the foundation for understanding and improving the present state of parenthood in America. The first part draws on demographic data, mostly from the United States, to identify the principal manifestations and sources of chaos in the lives of families living in economically developed societies. The second part introduces the bioecological model of human development. The third part moves from the model to experimental programs designed to counteract the chaos.

Most of the data, and most of the chaos for that matter, come from the United States. The chaos, however, is contagious, and is spreading to other societies, although as yet in much less violent form. Today's researchers on parenting are paying more attention to analyzing the developmental disarray of children than to the possible scientific bases and strategies for turning it around, and thereby actualizing untapped constructive potentials.

The Growing Chaos

Before we can turn the chaos around, we need to know what it is. In 1996 my Cornell colleagues and I published a volume documenting the marked changes that have taken place over the past four decades in the lives of children and youth growing up in economically-developed nations, particularly in the United States.1

Two main trends reinforce each other over time. The first reveals growing chaos in the lives of children, youth, and families. The second documents the consequences of this trend; namely, progressive decline in the competence and character of successive generations as they move into the 21st century.

Among the most prominent developmental trends are the following:

  1. Over the past two decades, systematic studies based on nationally-representative samples document increasing cynicism and disillusionment among American adolescents and youth manifested in a loss of faith in others, in the basic institutions of their society, and in themselves. For example, over a 12 year period beginning in the 1980s the percentage of U.S. high school seniors agreeing with the statement, "Most people can be trusted" fell by more than half from 35% to 15%.

  2. A complementary theme is increasing self-centeredness and disregard for the needs of others. Consider the change over time in response to the following item: "A man and a woman who decide to have and raise a child out-of-wedlock are 'doing their own thing and not affecting anyone else.'" One wonders what the picture is now.

  3. Ever greater numbers of American youth are becoming perpetrators and victims of crime. The levels for other countries --- Canada, Germany, England, and Japan --- are minimal or non-existent. The homicide rates for males aged 15-24 in the United States tripled from 12/100,000 in 1965 to 35/100,000 in 1990.

  4. More and more youth are spending their formative years in prison.

  5. Overall, rates of teen-age pregnancy and births are continuing to rise, albeit more slowly.

  6. Standardized measures of school achievement have been falling, even for students in the top 10% of the distribution.

These findings also indicate that the rising developmental disarray of children is the product of marked and continuing changes, taking place over the same time period, in the social institutions and informal structures that have greatest impact on the development of competence and character in the next generation. Among the most consequential of these social changes are the following:

  1. There has been a dramatic growth of single-parent families, whether through divorce or having a child without ever being married. For example, the percent of American children under 6 who are being raised by a single parent has doubled from 10% in 1970 to over 20% in the middle 1990s. As a result, the U.S. leads the developed world in the percent of children growing up in single-parent families and in teenage births. The critical problem here is whether there is a second parent figure present on a regular basis who not only cares for and engages in activities with the child, but also provides support, both material and emotional, to the single-parent mother.

  2. With more and more parents working full time, there has been a decline in the involvement of parents as active participants in and mentors of activities with children and youth. There is growing conflict between the demands of work and family.

  3. At the same time, the teen-age and adult models widely watched by children and youth on the media (TV, films, video games, CDs, and the internet) continue to emphasize commercialism, sexuality, substance abuse, and violence. The end result is a lack of positive adult models for internalizing standards of behavior and longer-term goals of achievement, and thereby an increasing number of autonomous peer groups bereft of adult guidance.

  4. Neighborhood ties among families have been eroding.

  5. A marked increase continues in the percentage of children and youth living in poverty, producing a widening gap between the rich and the poor.

  6. More and more of these trends are occurring at the same time, thereby increasing the pace, the scope, and the power of their developmentally disruptive effects.

Cross cultural comparisons are revealing. English speaking countries lead the world in teenage births, single parent families, and divorce in the following order: US, UK, Canada, Australia, Sweden, France, Germany, Japan. An explanation for this can be found in the observations by Alexis de Tocqueville in the 1830s. He noted that the young United States of America had two distinctive national characteristics. First, it was the most individualistic society in human history; second, it was also the most "volunteeristic." As I wrote some years ago: "We Americans are all the descendants of those who couldn't stand authority, and of those whom authority couldn't stand."

In his classic study Democracy in America, de Tocqueville pointed out that individualism had its roots in England. Hence, the "de Tocquville hypothesis" for our own times: namely that, after the United States, the societies showing the highest levels of social and developmental disarray will be other English-speaking countries.

The Forces Driving Human Development

There are many ways of knowing --- philosophy, literature, art, history. . . does science differ from these fields? Science is the only way of knowing in which you are obligated to try to prove yourself wrong. As Albert Einstein noted: "In science, more important than finding the right answers is to ask the right questions." And how does one find the right questions about how we develop as human beings?

Developmental Science in the Discovery Mode

Human development, or as some now prefer to call it --- developmental science --- is the scientific study of the conditions and processes shaping the biopsychological characteristics of human beings through the life course and across successive generations.

Here the principal aim is not the customary one of verifying hypotheses already formulated. It is a more extended process involving a series of progressively more differentiated formulations, with the results at each successive step setting the stage for the next round. The corresponding research designs must therefore be primarily generative rather than confirmatory versus disconfirming. Thus, the procedure is not the usual one of testing for statistical significance. Rather, the research designs must provide for carrying out an equally essential and necessarily prior stage of the scientific process: that of developing hypotheses of sufficient explanatory power and precision to warrant being subjected to empirical test. In short, we are dealing with science in the discovery mode rather than in the mode of verification.

At the same time, as in any scientific endeavor, it is essential that the successive formulations and the corresponding research designs be made explicit, and for this purpose it is necessary to have a systematic conceptual framework within which evolving formulations and designs can be classified and ordered in terms of their stage of scientific development in the discovery process.

For this dual purpose I have proposed the process-person-context-time framework (PPCT for short). Each of these four terms stands for a feature that has been used as a basis for investigating human development. For example; 1) from a historical perspective, one of the earliest elements employed for this purpose was a person characteristic --- the person's age; 2) environmental contexts such as social class and family structure did not enter the developmental research scene until the early 1900s; 3) the implications for human development of processes --- exchanges of energy, such as conditioning, and reinforcement (the forerunners of today's parent-child interaction first recognized during the 1930s); 4) investigations of developmental change through time over the life course and across successive generations are mainly a phenomenon of the last quarter century.

The PPCT framework does not constitute either a specific theoretical model or a corresponding research design. Rather, its purpose is to provide a taxonomy --- a system for classifying natural phenomena (from the Greek word taxis meaning "order") --- for identifying the defining properties of a particular theoretical system and its operational model.

With an appropriate taxonomy at our disposal, we are now in a position to examine the most recent reformulation in the discovery mode.

The Bioecological Model of Human Development

The basic structure and content of the bioecological model are defined in a series of propositions.2

Proposition 1

In order to develop --- intellectually, emotionally, socially, and morally --- a human being, whether child or adult, requires the same thing: active participation in progressively more complex, reciprocal interaction with persons, objects, and symbols in the individual's immediate environment. To be effective, the interaction must occur on a fairly regular basis over extended periods of time. Such enduring forms of interaction in the immediate environment are referred to as proximal processes. Proximal processes are posited as the primary engines of development.

Proposition 2

Proximal processes cannot structure, steer or sustain themselves. Their form, power, content, and direction vary systematically as a joint function of the characteristics of the developing person and of the environment --- both immediate and more remote --- in which the processes are taking place; the time through the life course and the historical period during which the person has lived; and the nature of the developmental outcome under consideration.

Note that the characteristics of the person actually appear twice in the bioecological model --- first as one of the four elements influencing the form, content, and direction of the proximal processes, and then again as the "developmental outcome"; that is, a quality of the developing person that emerges at a later point in time as the result of the mutually influencing effects of the four principal elements of the bioecological model. In sum, in that theoretical model, the characteristics of the person function both as an indirect producer and as a product of development.

Examples of Bioecological Research

How does the bioecological model fare when analyzed in a PPCT framework? Which elements are present, and how are they presumed to relate to each other? Two studies, each conducted some years ago, when analyzed in PPCT terms, come close to meeting the requirements of this theoretical model and its corresponding research design.

Drillien's Study

The first example dates from the late 1950s and early '60s. At that time, Cecil Mary Drillien, a physician and Professor of Child Life and Health at the University of Edinburgh in Scotland, carried out a seven-year longitudinal investigation of psychological development in two groups: 360 children of low birth weight, and a control group selected by "taking the next mature birth from the hospital admission list."3

Drillien's interest was two-fold; first, to analyze the impact of the quality of mother-infant interaction at age 2 on the frequency of problem behaviors observed in the infant at age 2, and again at age 4; second, to examine how this relationship varied as a joint function of the family's social class, and three levels of infants' birth weight --- those underweight by a pound or more, not more than one pound, and those of normal weight. Assessments of maternal responsiveness were based on observations in the home and interviews with the mother.

Drillien's measure of social class was a composite index that took into account not only parental income and education, but also the socioeconomic level of the neighborhood in which the family lived. The quality of interaction was assessed in terms of the extent to which the mother was responsive to changes in the state and behavior of the infant. Finally, the measure of the developmental outcome was based on the frequency of reported behavior disturbances, such as hyperactivity, overdependence, timidity, and negativism.

This research design includes all four elements of what is now called the bioecological model. Drillien's measure of maternal responsiveness closely approximates the definition in Proposition I of proximal process as the mechanism driving human development, with the further stipulation that, to be effective the process must occur "on a fairly regular basis over extended periods of time". To complete the picture, the remaining two elements are specified in Proposition II as characteristics of the person and of the environmental context, appear respectively in the form of the infant's birth weight and of social class. Finally, as previously noted, a characteristic of the person appears again, but now in the role of a developmental outcome,in this instance, the average number of problem behaviors exhibited both at age 2 and then again at age 4.

With theory and design in place, we turn to the results. In all three social class levels, the infants who had experienced low levels of maternal responsiveness at age two showed higher levels of problem behavior two years later, especially those youngsters growing up in the poorest environments. The proximal process (maternal responsiveness) markedly reduced the frequency of later problem behaviors, but the moderating factor was in a different quarter. Whereas maternal responsiveness made its greatest impact on those children growing up in the most disadvantaged environment, within that environment youngsters with normal birth weights benefited most from the developmental process.

Maternal responsiveness across time, a one-sided, two-category measure of proximal process, still emerged as an exceptionally powerful predictor of developmental outcome. In all instances, responsive maternal treatment reduced significantly the degree of behavioral disturbance exhibited by the child.

Herein lies the main justification for distinguishing between proximal process, on the one hand, and the environments in which the processes occur, on the other hand; namely, in accord with Proposition I, the former turn out to be an especially potent force influencing the developmental outcome (in this case, the frequency of problem behaviors two years later, when the children were 4 years old. Furthermore, as stipulated in Proposition II, the power of the Process varies systematically as a function of the environmental Context (i.e., social class) and of the characteristics of the Person (i.e., weight at birth).

However, one other key element of the bioecological model still remains to be considered. Proposition II also stipulates that the "form, power, content, and direction of proximal processes effecting development" also "vary systematically as a function of . . . the nature of the developmental outcomes under consideration."

Small and Luster's Study

Our next research example speaks to this issue. It also illustrates a "next stage" of developmental science in the discovery mode. Specifically, we have anticipated on theoretical grounds that the greater developmental impact of proximal processes on children growing up in poorer environments is to be expected only for outcomes reflecting developmental dysfunction. In this context, the term refers to the manifestation of difficulties in maintaining control and integration of behavior across a variety of situations. By contrast, competence refers to the manifestation and further development of knowledge and skills --- whether intellectual physical, emotional, or a combination of them (for example, learning how to care for an infant involves all three).

The theoretical expectation that proximal processes will differ in their developmental effects depending on the quality of the environment rests on the following basis. In deprived and disorganized environments, manifestations of dysfunction in children are likely to be both more frequent and more severe, with the result that they attract more attention and involvement from parents, whereas in advantaged and more stable environments, such manifestations are less intense, and parents are more likely to be attracted by and respond to gratifying signs of their children's developmental progress.

In addition, most parents, have the capacity and the motivation to respond to the immediate physical and psychological needs of their children. The situation is rather different, however, with respect to enabling their children to acquire new knowledge and skill. In this domain, either the parents must themselves possess the desired knowledge and skill, or they must have access to resources outside the family that can provide their children with the experiences needed to develop competence.

Taken together, the foregoing considerations lead to a working hypothesis regarding the differential impact of proximal processes as a joint function of the quality of the environment in terms of available resources, on the one hand, and, on the other, the nature of the outcome in terms of competence versus dysfunction.

Some indication of the validity of this hypothesis is provided by the results of an analysis depicting the differential effects of parental monitoring on school achievement for high school students living in the three most common family structures found in the total sample of over 4000 cases.4 The sample is further stratified by two levels of mother's education, with completion of high school as the dividing point. Parental monitoring refers to the effort by parents to keep informed about and set limits on their children's activities outside the home.

Once again, the results reveal that the effects of proximal processes are more powerful than those of the environmental contexts in which they occur. In this instance, however, the impact of the proximal process is greatest in what emerges as the most advantaged ecological niche --- families with two biological parents in which the mother has had some education beyond high school. In single parent and stepfamilies the same degree of active effort yields a somewhat smaller result. Thus, in this case, for pupils who are not doing so well in school, parental monitoring can apparently accomplish a great deal by insuring stability of time and place so that some learning can occur. But superior school achievement would clearly require in addition high levels of motivation, focused attention, prior knowledge, and, especially, actually working with the material to be learned, all qualities that stability of time and place by themselves cannot provide.

Within each family structure, parental monitoring exerted a more powerful effect on the school achievement of girls than of boys, a result that is paralleled by corresponding differences in average GPA for the two sexes. In each of the three family structures, girls received higher grades than boys, with the difference being most pronounced in two-parent households and lowest in single-mother families.

A distinctive feature of the pattern for girls is a marked flattening of the curve of scholastic achievement, especially for daughters of single-parent mothers. This result suggests that, in each of the three family structure, better educated mothers may be pushing their already successful daughters too hard to the point where conformity to maternal control no longer brings educational returns, particularly when the mother is the only parent.

An analysis of data on students whose mothers had no more than a high school education showed a similar general pattern, but the effects were less pronounced. The influence of monitoring was appreciably weaker, and its greater benefit to girls was also reduced. Nevertheless, girls with less educated mothers both in single-parent and in stepfamilies still had higher GPA scores than boys.

Nature-Nurture from the Bioecological Perspective

Today, a growing body of research claims strong evidence for the view that individual and group differences in a wide range of developmental outcomes are mainly driven by genetic endowment.5,6 In response, Bronfenbrenner and Ceci have proposed an empirically testable theoretical model that (a) goes beyond and qualifies the established behavioral genetics paradigm by allowing for non-additive synergistic effects, direct measures of the environment, and mechanisms of organism-environment interaction (namely, proximal processes) through which genotypes are transformed into phenotypes; (b) hypothesizes that estimates of heritability increase markedly with the magnitude of proximal processes; (c) demonstrates that heritability measures the proportion of variation in individual differences attributable only to actualized potential, with the degree of nonactualized potential remaining unknown; and (d) proposes that, by enhancing proximal processes, it is possible to increase the extent of actualized potentials for reducing developmental dysfunction, and increasing developmental competence. This alternative formulation still awaits a rigorous test.

Applying the Bioecological Model

The examples considered thus far are essentially "experiments of nature"; that is, they show how development is influenced by variation in the elements of the bioecological model occurring in already existing social conditions. But they tell us nothing about whether, to what extent, or how these elements and their combinations, can be changed. This limitation applies particularly to the most consequential component of the bioecological model --- proximal processes. The most effective way to answer this question would be to conduct an experiment, with subjects randomly assigned to different experimental conditions, including a control group.

In fact, such an experiment has been carried out. In 1978, Marianne Riksen-Walraven, a developmental psychologist in the Dutch city of Nijmegen, conducted an experiment with a sample of 100 9-month-old infants and their mothers.7 Because existing findings indicated that this was the group in greatest need, "all subjects came from working class families." In the research design, mother-infant pairs were randomly assigned to one of three groups. Those who ended up in what Riksen-Walraven called the "stimulation group" (Group 1) were given a "Workbook for Parents," with drawings to match, emphasizing the importance of mothers' providing their infants with a variety of experiences that captured the baby's attention; for example, "pointing to and naming objects and persons," and "speaking a lot to their infants."

By contrast, the workbook for mothers in the "responsiveness" group (Group 2) stressed the idea that "the infant learns most from the effects of its own behavior."8 Accordingly, caregivers were advised not to direct the child's activities too much, but to give the child opportunity to find out things for himself, and then to respond to the child's initiatives.

Finally, mothers in the third experimental group (Group 3) were given pages from both workbooks, in effect recommending that the mothers use both strategies.

How did the three groups come out in the follow-up conducted three months later? Which group did the best, and which did the worst?

All three experimental groups were influenced by the workbooks they were given. However, it was the children in the "responsiveness" group (Group 2) who showed the strongest and most pervasive effects on laboratory measures of young children's cognitive development that were administered at the end of the experiment. Specifically, they exhibited the highest gains on measures of exploratory behavior, were more likely to prefer a novel object to one that was already familiar, and obtained higher scores on a learning task. The "stimulation" group placed second; and the one recommending both "stimulation and "responsiveness" did least well.

Why? What can be inferred from using both strategies together? The author does not address this question, but one possible explanation is that the mothers were faced with having to make a decision about what to do when, thus interrupting a free flow of interaction.

The final analysis revealed marked differences in maternal behavior corresponding to the advice and examples presented in the "Workbook for Parents" that they had been given. This set of results suggests the possibility of designing home-based intervention programs that could operate effectively at much lower cost and reach many more families in need than those requiring on-site professional staff on a regular basis.

A key question, however, is how long the effects can last. Fortunately, the original investigator, together with a colleague, has provided an answer in a longitudinal follow-up study of the same families when the children were 7, 10, and 12 years of age, using measures based in teachers' ratings.9 The question is well advised, because at these later ages only the girls from the original Responsiveness group showed any effects of the original experimental intervention. Specifically, at all three later ages they were rated by teachers as more "competent and skillful," "curious and exploring," "resourceful in initiating activities," "better able to handle stressful situations," "less dependent on adults for help," and "less anxious." In addition, by age 12, these girls were described by their teachers as more "attractive, interesting and energetic children."

The authors offer the following speculations "of why the Responsiveness program was particularly effective for girls and not for boys:"

We suggest that different parental attitudes towards competence striving in boys and girls may have contributed to this finding. Striving for competence, independence, and self-reliance are part of the traditional masculine stereotype, which was still current in the early seventies when our intervention started. . . This means that initially competent boys will have more opportunities to experience themselves as effective agents than initially competent girls. . . . It thus seems possible that competence motivation can be maintain itself in young girls only when it is at an 'extra' high level and only when the parents are 'extra' willing to accept their daughters' autonomy and independent exploration.

In sum, from the perspective of the bioecological model, we have evidence for the power of an experimentally-induced proximal process in furthering young girls' psychological development. This result was accomplished by changing the belief systems of mothers thereby leading them to provide a different kind of experience for their young daughters over an extended period of time. Thus, taken as a whole, the findings encompass all four defining properties of the bioecological model and its corresponding PPCT design, with developmental context deliberately limited to working class families as the group most in need of assistance. The fact that the long-range effects of the intervention were limited to girls poses an unanswered question regarding the nature of proximal processes that might achieve similar developmental gains for boys.


We Americans have yet to confront the reality that the growing chaos in the lives of our children, youth, and families today simultaneously pervades too many of the principal settings in which we live our daily lives --- in the family, health care systems, child care arrangements, peer groups, schools, neighborhoods, the workplace, and means of transportation and communication between them.

To be sure, these also are the settings in which our society has concentrated resources and efforts to reverse the mounting developmental disarray. Even though the United States experienced an economic upswing in the late 1990s, the sparse bits of more recent demographic data I have been able to obtain give little indication of a true and lasting turnaround. The rising trend of chaos and its consequences also extend to other spheres of our society. Not long ago, one of our nation's leading corporate executives gave a major lecture at Cornell's Graduate School of Management. His title: "Growing Chaos in America's Corporate Enterprises." He said that no sooner is a new production policy implemented after weeks of planning and testing, than an order comes down from above "to scrap the whole thing" because the policy has been changed.

Is there a known strategy that can reverse disruptive changes that are so powerful and widespread? Yes there is, and it has a long history. But I know of only one instance in which such a strategy was conceived and carried out by developmental scientists. The strategy was based on theory, refined in an "experiment of nature" and was applied in an "experiment by design." The central figure in this remarkable achievement was the developmental psychologist Lev Semyonovitch Vygotsky. The central idea underlying the entire enterprise was Vygotky's concept of the "transforming experiment." By this he meant an experiment that restructures the environment to produce a new configuration that activates the previously unrealized developmental potential of the persons living in that environment.

What was the nature of Vygotsky's experiment and of its findings? Alexander Romanovich Luria, one of Vygotsky's best known students, tells the story in his autobiography. The time was the early 1930s.

We conceived the idea of carrying out the first far-reaching study of intellectual functions . . . By taking advantage of the rapid cultural changes that were then in progress in remote parts of our country, we hoped to trace the changes in thought processes that are brought about by technological change. . . At that time, many of our rural areas were undergoing rapid change with the advent of collectivization and the mechanization of agriculture.

The basic research design took advantage of the fact that the process of modernization had not been introduced in all areas of the Soviet Union at the same time. As a result, it became possible to carry out a comparison of cognitive functioning in communities differing in their degree of exposure to social change. Vygotsky died of tuberculosis before this extraordinary investigation was completed.

The following is Luria's succinct summary of the findings:10

Our data indicate that decisive changes can occur in going from graphic and functional --- concrete and practical --- methods of thinking to much more theoretical modes of thought brought about by changes in social conditions, in this instance by the socialist transformation of an entire culture.

The publication of the study in the Soviet Union was held up for more than three decades. The reasons for the delay are perceptively described by Michael Cole in his preface to the American edition of Luria's book: "The status of national minorities has long been a sensitive issue in the USSR, [not unlike the issue of ethnic minorities in the United States]. It was all well and good to show that uneducated, traditional peasants quickly learned the modes of thought characteristics of industrialized socialist peoples, but it was definitely not acceptable to say anything that could be interpreted as negative about these people at a time when their participation in national life was still so tenuous."11

Transforming experiments have also been carried out in the United States, but regrettably their developmental effects have never been investigated systematically. Perhaps one of the most successful was the G.I. Bill, which gave new hope and a new life to a whole generation of World War II veterans and their families. Alas, the same deserved legacy was not bestowed on their comrades-in-arms in the wars that followed.

A second nominee is Head Start. But from what I know both from looking at its budgets, mounting bureaucratic controls, and from personal experience as an external member of the Head Start Parents' Policy Committee, the prospects for the future are hardly rosy. Head Start parents are today drawing on their own meager resources to continue some of the programs that are needed most, and they take time off from jobs (when they have them) to help fellow families in emergencies because of illness or the desperate need for child care.

Such heroic acts are signals to the rest of American society. They sound a call for our own transforming experiment, one that can draw on the deepest sources of our national strength. As yet, this call, and many others, are not being heard, either by our scientists or by our citizens. We do not heed the immortal words of John Donne: "Do not ask for whom the bell tolls; it tolls for thee."


1 Bronfenbrenner, U., McClelland, P., Wethington, E., Moen, P. & Ceci, S.J. (1996) The state of Americans: This generation and the next. New York: The Free Press.

2 Bronfenbrenner & Morris (1998) in (Series Editor, Damon, W.) Lerner, R.M. (Ed.) Handbook of Child Psychology, Fifth Edition, Volume I: Theory, pp. 993-1028.

3 Cecil Mary Drillien (1957, p. 29, 1964)

4 Small, Steven and Luster, Thomas (1990)

5 Plomin, 1993

6 Scarr, 1992

7 Riksen-Walraven (1978) p. 111).

8 Ibid, p. 113.

9 Riksen-Walraven & van Aken (1997)

10 Luria, A.R. (1976) p. vi.

11 Ibid, p. xiv.

Copyright © 1999 Urie Bronfenbrenner.

stars icon For technical assistance:
Library Technology Group
General Library System
University of Wisconsin-Madison