Why Should We Invest in Adolescents?

Paper prepared for the Conference on Comprehensive Health of Adolescents and Youth in Latin America and the Caribbean, July 9-12, 1996 and Pan American Health, W. K. Kellogg Foundation, Washington, DC, 1998.

The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.

---

Contents

Framework for Understanding Initiatives toward Adolescents

Circumstances and Needs of Latin American and Caribbean Youth

What Payoffs Can We Expect from Investing in Activities that Promote Adolescent Health?

Recommendations for Shaping and Targeting Investments in Adolescents

Summary and Conclusions

References

Notes

Tables

---

Framework for Understanding Initiatives toward Adolescents

Without the realistic hope of getting ahead economically, there is little incentive for youth to invest in education or protect themselves from some of the less healthy habits they may acquire during adolescence.

Too often in the past, public policy has either ignored adolescents or focused on them only when they behave in ways that trouble their elders. Compared to very young children and to the elderly, adolescents suffer from few life-threatening conditions. The formation in adolescence of certain health habits with long-term negative consequences (such as smoking tobacco products, use of other addictive substances, or sexual activity without protection from STD and AIDS) often does not produce morbidity or mortality in adolescence itself. Rather the effects, and the costs, develop over a lifetime. Thus, when societies face decisions about where to invest significant health and other supportive resources, attention to adolescents often receives short shrift, despite the fact that after early infancy, adolescence is the period of greatest vulnerability until one gets to the diseases of old age.

Need for Future Health and Productivity

Adolescents aged 10-19 are a very large and significant proportion of the population of Central and South American and Caribbean countries, ranging from 17 percent in Uruguay up to almost 26 percent in El Salvador. For all of Latin America the percentage is 21.7 percent, with figures lower than 20 percent only in Argentina, Chile and Uruguay. This compares to the much lower 13.7 percent for North America (Knaul and Fl&#oacute;rez 1996, Table 2). To a large degree, the future economic development of these countries depends on having increasing proportions of the population who are reasonably well-educated, healthy, and economically productive. Being both well-educated and healthy contributes to potential economic productivity, but of course does not guarantee it. It is also essential that countries create increasingly broad and rewarding economic opportunities. There is also an important interaction between economic opportunity and the readiness of today's youth to take advantage of it. Without the realistic hope of getting ahead economically, there is little incentive for youth to invest in education or protect themselves from some of the less healthy habits they may acquire during adolescence. But without the expectation that there will be a qualified workforce to fill newly-created jobs, potential investors may be reluctant to make the necessary commitments to economic development. To the extent that a country's youth of today and tomorrow are not prepared for that future (and many are not, as will be documented below), hopes for the country's economic future become increasingly dim.

Problem-Focused Versus Holistic Programming

Another critical issue has to do with the programs that do exist to serve adolescents, in Latin American countries and also in other parts of the world. The types of attention that adolescents usually do receive, when they receive any, too often is focused on specific problem behaviors once the behaviors are fairly well-entrenched. Prevention programs are relatively scarce; what happens instead can be called "tertiary attention"—trying to fix something after it is badly broken (Barker and Fuentes 1995). Even more significant, these tertiary efforts usually focus on single issues such as early childbearing, or substance abuse, or criminal involvement. They do not take a holistic approach to youth, their families, their environments, and the overall context in which their behavior occurs. As we will see below, these single-focus, tertiary types of attention have not demonstrated much significant capacity to change the lives of adolescents. They are, however, usually quite expensive on a per-client basis. Later, this paper will examine the evidence and look at the breadth of the growing consensus that interventions to promote adolescent health are most effective when they use a holistic framework.

Framework for Working with Adolescents

The remainder of this introductory section will focus on articulating the different elements of such a framework. In its final section, this paper will return to the framework to examine the practical issues and challenges involved in conducting work within its context.

Theories of Youth Development and Their Implications

Adolescence was once viewed as a life stage that, by its very nature, involved serious conflict and upheaval as individuals tried to sever their connections with childhood dependence and struggled to achieve an independent adult identity (Blos 1962; Freud 1958). The troubles of adolescents were seen as typical or to be expected during this life stage, rather than as a sign that something was seriously wrong. However, more recent thinking sees much less necessary difficulty with the process and much more continuity between the child that was, through the adolescent that is, to the adult that will be. This view, expressed by Offer and colleagues (Offer et al. 1981; Offer 1987) and others, is that normal adolescents negotiate this period of life transition with relatively little major disruption or sustained high risk behavior, maintaining and developing their own identity and their relationships with parents as well as adding elements of identity and building new extra-familial relationships and skills. Given this view of "normal" adolescence, those teenagers who do experience major disruptions and who do persistently engage in problem behaviors are in trouble now, and have a significantly greater chance of being in trouble later in life (Hamburg and Takanishi 1989). Therefore, successful interventions with these youth are likely to have important payoffs in terms of future health problems prevented and future satisfying and productive lifetimes promoted.

Developmental Tasks of Adolescence

As with all stages of life, adolescence involves some key tasks that build on earlier successful development. As adolescence is the transition period between childhood and adulthood, the tasks of this period are all related to making that transition. Adolescence is the period in which individuals face the tasks of establishing a satisfying self-identity and interpersonal bonds beyond the family, including partnering; learning to handle growing sexual maturity in a responsible manner; and developing the capacity for economic viability, including education, skills, attitudes, habits. Adolescents' family, peers, neighborhood environment, school, and other associations can help them complete these tasks, or can pose significant barriers which many youth will not be able to overcome on their own. We can begin to make progress in helping adolescents once we can begin to understand how even the most undesirable of youth behaviors usually represents either: 1) the attempts of adolescents to complete these developmental tasks; 2) ambivalence about whether they want to move on to adulthood or to stay a child; or 3) the consequences of their perception that they may never complete the tasks successfully.

Addressing Symptoms, or Underlying Conditions

As has already been noted, most programs serving adolescents focus on specific conditions and usually do not intervene until the conditions have reached the level of "problems." Thus we have programs dealing with teenage childbearing, programs dealing with failure to complete secondary school, programs dealing with substance abuse among youth, and programs trying to reduce youth violence. To a large extent this compartmentalized pattern of services reflects the narrow focus of many government agencies. However, a great deal of research on the factors contributing to the development of these problems reveals that they have common antecedents—the same underlying conditions are identified time and again.

In a recent summary of research on risk factors, Catalano and Hawkins (1995) identify the following as common antecedents for most or all of substance abuse, delinquency, teenage pregnancy, school dropout and violence: extreme economic deprivation, family conflict, family history of the problem behavior, and family management problems. Further, substance abuse, delinquency and violence all share neighborhood characteristics that suggest the neighborhood offers particular opportunities to engage in problem behaviors, and little to help youth resist them, namely: community laws and norms favorable toward drug use and crime; availability of firearms, friends who engage in the problem behaviors, parental attitudes favorable to the problem behavior, and low neighborhood attachment and community disorganization. Under these circumstances, youth striving to achieve identity, skills, and a livelihood may have ready access to problem activities (from the society's point of view) and little access to positive ones.

Too often, public policies and programs address youth problems as if they happen in a vacuum. A more apt metaphor for youth problem behavior appears to be a volcano. The volcano has a seething, molten inner core, and a number of potential fissures through which the pressure building up at the core may be relieved. One fissure may be risky sexual behavior; another may be involvement with drugs; a third may be criminal behavior; a fourth may be school dropout. If we focus on the symptom emerging from a particular fissure and try to fix it, say by plugging the fixture, the underlying pressure from the life conditions of high risk youth will simply find another way to escape. Perhaps this will be as another problem behavior. Perhaps, with proper guidance and intervention to help youth achieve developmental tasks in a constructive manner, it could be legitimate entrepreneurial activity or community service. The important point is that, to change the fate of youth in these circumstances, we have to find ways to reduce the pressure or turn down the heat at the base of the volcano. This means addressing the underlying causes or conditions of adolescent lives, at the same time we think about how to help youth toward successful completion of their fundamental developmental tasks.

Risk and Resiliency

Just as there are factors in a youth's environment and background that condition the probability that the youth will get into trouble, there are also factors that may be able to protect youth from their influence. The idea of protective factors grows out of work that focuses on children who seem to do well despite living in circumstances that expose them to considerable risk (Cowen and Work 1988). These children have been called "resilient," "invulnerable," "competent," and "stress-resistant," among other appellations (Garmezy 1983, 1987; Rutter 1987; Werner 1986, 1989; Werner and Smith 1982). Much recent research has focused on trying to identify such factors, and to learn how they work (that is, whether they are independent, positive factors in their own right, or whether they have effects primarily when there are risk factors that need to be countered or alleviated).

Individual, familial, and environmental protective factors have been identified to date, many of which are found to operate as independent positive factors whether or not a youth also experiences a high level of risk factors. Grossman et al. (1992) summarize these factors. Protective individual (personality) factors include self-esteem (Garmezy 1983; Murphy and Moriarity 1976; Rutter 1979; Werner and Smith 1982), and internal locus of control (feeling confident that your own efforts will produce desired effects (Garmezy 1987; Werner 1986). Protective familial factors include the absence of marital discord (Garmezy 1987; Rutter 1987), family cohesion (Felner at al. 1985), and a good relationship with at least one parent (Campbell 1987; Hauser et al. 1985; Kwakman et al. 1988; Robertson and Simons 1989; Rutter 1979). Having an important relationship with a significant adult figure other than one's parent(s) is a protective factor in the social environment beyond the family (Garmezy 1983; Murphy and Moriarity 1976; Rutter 1979; Werner and Smith 1982). In some studies (e.g., Werner 1989; Werner and Smith 1982), the need for protective factors appears to increase as the risk factors to which a child or youth is exposed increase. Thus more youth will successfully negotiate circumstances of low risk without protective factors than will successfully negotiate circumstances containing many risk factors unless they also have some protective factors working for them.

The Service Implications of Youth Development Theories

The view of adolescence just presented has important implications for how we should work with adolescents. First, it is critically important to develop coordination among different parts of the service and support system so the needs of adolescents can be addressed in a holistic and coordinated way. Even if health is the major focus of a program, it is still necessary to address the needs of youth to earn money, help their family or learn how to deal with its difficulties, and feel confident in their abilities, or these things may get in the way of youth being able to follow appropriate health regimens.

Given the potential importance of protective factors, many interventions with high risk youth try to supply or strengthen them. Many programs work on improving youth self-esteem and sense of self-efficacy, regardless of what other aspects of youth experience they are trying to influence. Programs also try to assess the problems that may be affecting adolescents' families, and offer assistance either directly to the family or to the youths as individuals in coping with their families' problems. When their family is homeless, or a parent is violent or a drug user, a youth will probably not have the ability to concentrate on schoolwork or keep up their sense of self-efficacy as long as the family problems persist, or until the youth gets some help in protecting him- or herself from negative family influences. Many programs also try to offer sympathetic and supportive adults with whom youth can bond, in order to supply this relationship as an additional protective factor. This is especially important when familial problems and difficulties with parents are among the chief risk factors in the lives of youth.

---

Circumstances and Needs
of Latin American and Caribbean Youth¹

Teenage sexuality, pregnancy, and childbearing present major health issues to Latin and Caribbean countries, as well as to many developed countries such as the United States.

This section briefly summarizes the conditions of Latin American and Caribbean youth, to the extent that available statistics make this possible. It looks at education, employment, health and reproductive health status, and substance abuse information.

Education

Educational attainment of youth varies widely among the countries of Latin America and the Caribbean basin, while within-country variations between urban and rural areas and among income groups are equally striking. Data are not available from every country. Eleven Latin American countries report data by income strata about the proportion of urban youth ages 20-24 who are not enrolled in school and who have less than 10 years of completed education; seven of the same countries also report data for their rural areas (CEPAL 1994, Table 47). For urban youth, the proportion achieving at least 10 years of education ranges from a high of almost 80 percent in Chile to a low of only 46 percent in Honduras. However, class differences loom large in all of the reporting countries. Chile and Venezuela have the smallest (but still substantial) class differences, with 92 percent of the highest income quartile and 62 percent of the lowest income quartile having this much education among Chile's urban youth; in Venezuela, 74 percent of urban youth in the highest and 43 percent in the lowest income quartiles have completed at least 10 years of education. Mexico reports the largest class differences for urban youth educational attainment, with 80 percent of youth in the upper income quartile completing at least 10 years of school compared to only 18 percent of urban youth in the lowest income quartile.

Parallel statistics on rural youth show much smaller between-class differences, but are much lower than the figures for urban areas in all countries reporting. Panama reports the best figures, and still only has 41 percent of rural youth completing at least 10 years of education. Figures from other countries range from 26 percent (of rural youth in Costa Rica) down to 14 percent (of rural youth in Honduras).

Many indicators from around the world suggest that the way a country treats the educational opportunities of girls is a key to modernization, and also a key to avoidance of many problems specifically related to youth. When girls have opportunities to be something in life in addition to childbearers and domestic laborers, many other things also change, including family size and child spacing, and child and maternal health. Singh and Wulf (1990, Table 3.1) report widely discrepant performance of different Latin American and Caribbean countries in educating girls up to the standard of 10 or more years of schooling completed. The highest proportion is 55 percent in Trinidad and Tobago, which has a serious commitment to promoting gender equality. The next highest drops dramatically to 31 percent in Ecuador and 21 percent in Peru. The lowest proportions are for Guatemala and the Dominican Republic, at 6 and 8 percent, respectively. However, the Dominican Republic and Guatemala should not be thought of as equivalent, since 54 percent of girls ages 15-19 have 7 to 9 years of schooling in the Dominican Republic, compared to only 13 percent in Guatemala.

Employment and Unemployment

Youth employment may be a positive thing, especially for older youth. It is particularly critical for those beyond school age, for whom the inability to find a job, or to find a decent one, is usually the issue. For younger adolescents, the issue is more that their employment, which may be essential in contributing to family income, interferes with their schooling and therefore with their potential to get better jobs as they reach adulthood.

For the most part, CEPAL (1994, Tables 31 and 32) reports that fewer than 10 percent of indigent and poor but non-indigent urban children ages 12 to 14 work in most Latin American countries (the corresponding figures for non-poor urban youth are under 5 percent). In two countries reporting data, however, Brazil and Paraguay, 18 and 23 percent of indigent urban youth in this very young age group are employed. Chile is the lowest, with only 2 percent of indigent and poor non-indigent urban 12-14 year olds working.

In most of the same countries, the picture changes dramatically for 15 to 17 year olds. For this age group, only Bolivia, Chile and Venezuela report employment rates at or under 15 percent among indigent urban youth. Many other countries report one-fourth, one-third, or even one-half of indigent urban youth in this age range are employed. In most of the same countries, the proportions employed do not drop substantially for the poor non-indigent urban 15 to 17 year olds, and in some countries they even go up a little. In Argentina, Bolivia, Brazil and Venezuela, non-poor urban youth ages 15-17 are at least as likely to work as those who are indigent. If work patterns of youth in developed Latin countries parallel youth work patterns in the United States, it is likely that the jobs being held by the more economically advantaged youth will actually contribute to future employability and earnings potential, whereas the types of jobs held by indigent and poor youth will contribute little to developing their skills and will be more likely to interfere with school completion than is true for better-off youth. Within a given country, rural youth of any age are more likely to be employed than their urban counterparts, regardless of the income class of the latter (CEPAL 1994, Table 44).

On the unemployment side, urban youth ages 15-24 are about twice as likely as the total population to want or need jobs but be unable to find them. This finding is consistent across many Latin countries, and holds true for both males and females. Youth unemployment as a percentage of all unemployment runs between 36 percent and 66 percent, with most countries falling in the 40 to 53 percent range (CEPAL, 1994, Table 12).

A high proportion of youth neither work, look for work, nor attend school, with the consequence that they miss opportunities for both education and training. In 1992, among urban households from the lowest income quartile, 25 percent of youths ages 13 to 17 in Honduras and Argentina neither worked nor attended school. In Brazil, Costa Rica, Uruguay and Mexico the figure was around 20 percent, and was about 16 percent in Colombia and Venezuela (CEPAL 1994, Table 45).

This group of out-of-school youth who are not in paying jobs includes many whose parents are poor or indigent, who themselves are poor or indigent, and whose children are likely to be poor or indigent. There is considerable likelihood of criminal involvement among this group of youth. Also, it contains many young women who are kept at home to do unpaid household labor, and are thus unable to attend school. Little is known about the longer-run effects of unpaid work in the home on the relatively high proportion of Latin American young women who perform it. But it certainly must impede their ability to acquire the skills and experience that will let them obtain significantly better jobs in the future (Fl&#oacute;rez, Knaul and M&#eacute;ndez 1994; Knaul 1995; Knaul and Fl&#oacute;rez 1996).

Health and Adolescent Pregnancy

The health problems of adolescence and youth in Latin America and the Caribbean tend to be very different from those of childhood, and are associated with the major organic and psychosocial changes that are part of adolescent development (OPS 1990). The developing countries experience significant differences in mortality rates and causes among adolescents. Infectious diseases (diarrhea, influenza, pneumonia) still stand among the five major causes of death for 10 to 14 year olds in countries like Guatemala, but in countries like Colombia that are closer to the other end of an epidemiologic transition, accidents and violence are more important as causes of death among adolescents. In some countries, violence due to police action and war is also a leading cause of death (OPS 1990).

Another issue in adolescent health is the prevalence of sexually transmitted diseases and HIV/AIDS. The highest incidence of STD is found in the age group of 20 to 24 year olds, followed by those from 15 to 19 and then those 25-29 years of age. At least half of those infected with HIV are 24 years old or younger.

Teenage sexuality, pregnancy, and childbearing present major health issues to Latin and Caribbean countries, as well as to many developed countries such as the United States (European nations are considerably better at preventing unwanted pregnancies and childbearing than is the United States, Jones et al. 1985.) Early sexual initiation and childbearing are associated with poor nutrition, poor or non-existent prenatal care, premature delivery, complicated childbirth, low birthweight babies and a higher proportion of babies born with additional complications, poorer infant outcomes, and the probability of less adequate parenting (Hayes 1987). The health risks to young women from illegal abortions are also great. Estimates for Latin American countries suggest that perhaps as many as 4 in 10 pregnancies end in abortion, going as high as 6 in 10 for Chile, and as low as 2 in 10 for Mexico (Singh and Wulf 1994). To the extent that teen pregnancies result from sexual violence or abuse, they are likely to have other long-term psychological and physical health consequences that affect demand on health services (Heise, Pitanguy and Germain 1994).

Substance Abuse and Other High-risk Behaviors

There are very few sources of country-level data reflecting the risk-taking behavior of Latin American and Caribbean youth. Those sources that do exist vary in the age range of youth studied and the ways that questions were asked. The older the age range, the more likely it is that studies find higher proportions of youth who have tried particular risky behaviors, or who habitually use addictive substances. Smoking is probably the behavior most common among youth of all ages, and most likely to affect health because of its prevalence (OPS 1990). Significant proportions of male youth smoke cigarettes (e.g., 57 percent of 15 to 19 year olds in Peru, and 41 percent in Cuba, compared to 28 and 32 percent in the United States and Canada, respectively). In Mexico, 17 percent of an even younger age cohort, the 11 to 15 year old males, were smokers. In the United States and Canada, female youth rates of smoking are virtually the same as for males, whereas in the Latin countries studied, female youth were significantly less likely to smoke (although rates were still high, 40 percent in Peru and 28 percent in Cuba). In Mexico there was little difference in male and female rates among 11 to 15 year olds.

---

What Payoffs Can We Expect from Investing in Activities that Promote Adolescent Health?

The evidence is very strong that the successful programs, even though nominally single-focus, in reality treat children and youth (and their families) holistically, start young, offer many enrichment, growth and development activities, and stick with youth for a long period of time.

One way to think about the importance of attending to adolescent health issues, and attending to them in a holistic and comprehensive way, is to think about what the consequences will be if we do not make these investments. No studies of which I am aware try to assess the personal or societal costs of adolescent risk behaviors for Latin American or Caribbean youth. However, several such attempts have been made in the United States, with respect to a variety of topics including teenage pregnancy and childbearing, educational completion and juveniles likely to become career criminals. I describe some of these studies in this section, with two purposes in mind. The first purpose is to give the reader an idea about how some analysts have tried to conceptualize the ideas of investment, payoff, costs, and social and personal consequences. There are a number of approaches, and there may be reasons for choosing one or the other of them depending on what you are trying to do. The second purpose is to give some concrete examples of analyses that may stimulate the reader to think about how similar investigations could be undertaken in their own countries. Toward the end of this section I describe the results from several studies that examined the effects of holistic, integrated service and support systems designed to help high risk youth.

Statements about costs that could have been prevented with appropriate interventions make very powerful arguments. For instance, the Carnegie Council on Adolescent Development (1993) recently presented some telling calculations for the United States:

• Each year's class of high school dropouts will, over their lifetime, cost the nation $260 billion in lost earnings and foregone taxes.
  
  
• Over a lifetime, the average high school dropout will earn $230,000 less than a high school graduate and contribute $70,000 less in taxes.
  
  
• Each added year of secondary education reduces the probability of public welfare dependency in adulthood by 35 percent (with associated reductions in public costs).
  
  
• Each year, the U.S. spends roughly $20 billion in payments for income maintenance, health care, and nutrition to support families begun by teenagers.
  
  

When thinking about justifying investments in public support and services to adolescents, there is some pressure to try to quantify the results, and also to monetarize them if possible. People usually approach the job of justifying investments by asking, "What is the cost of not helping high risk youth—that is, suppose we did nothing and the youth went ahead and pursued high risk behaviors, what would be the consequences? And, are those consequences sufficiently disastrous, for a youth, for a community, or for society as a whole, that we should seriously consider putting some public resources (money, programs, attention) into trying to prevent them and into promoting their opposite?"

Several components of cost have been investigated, including:

• Things of value to the individual that he or she is less likely to attain if he or she indulges in risky behavior—the most obvious monetary example is income that might have been earned if the individual had completed high school rather than dropping out, or had lived rather than died of a drug overdose, or had not had a baby at the age of 14. "Years of life foregone" is a quantified but not monetarized measure often used to assess the seriousness of various causes of death.
  
  
• Things of value to society that the individual is less likely to produce if he or she indulges in risky behavior—here the most obvious example is the taxes of several varieties that the individual will not pay because he or she is in prison, or dead, or stuck in a dead-end low-paying job in the informal sector.
  
  
• Things that cost public money, that are more likely to be needed by the individual or by the society if he or she indulges in risky behavior. These include health care, social services, special educational services, income assistance, police activities, prison cells, etc. These can usually be both quantified and monetarized, since we usually can develop estimates of what they cost, both by the unit (e.g., one doctor's visit or one hospital stay for a premature delivery with complications), and for a whole city, state, department, or country. Some investigations may also ask how much private money is being spent on the same types of services, but usually this type of information is harder to get.
  
  

Research Studies

Of course, different cultures will value each of these potential cost components differently, whether monetarized or not. So every time a researcher begins to think about how to justify making investments in adolescents it will be necessary to think carefully about the personal and societal outcomes that are most valued in that culture, the personal and societal outcomes that are most unwanted, and the capacity of available data to support measurement of the elements you care about the most. Any given piece of research such as the ones described below can only offer suggestions and guidelines which other researchers must adapt to their own circumstances.

Teenage Childbearing

During the early 1980s, several pioneering attempts were made in the United States to estimate various aspects of the costs of teenage childbearing. Among the most sophisticated were Wertheimer and Moore (1982), who used dynamic simulation computer modeling to estimate lost earnings, lost work years, lost taxes, and probable use of public benefits such as welfare (Aid to Families with Dependent Children) and public medical care (Medicaid); and SRI International (1979), who developed a 20-year projection of the public welfare, medical, housing and social services costs of a family begun by a first birth to a teenager. Both studies controlled for the age at first birth, and assumed higher costs the younger the mother at first birth².

Burt (1985, 1986) took the findings from these studies and developed two simple cost-estimating methods that could be conducted in any jurisdiction of interest (county, state, or for the whole country). The first method calculates the money spent through the three biggest public assistance programs in the United States during a single calendar year that went to families begun by a teen first birth.

The second method follows the SRI International model, and estimates the cost to the public over a 20-year period, in welfare, medical care, housing, and social services, of each family begun by a birth to a teen younger than 15, 15 to 17, and 18 to 19. To learn the total public cost for an entire calendar year's cohort of first births to mothers 19 or younger, these calculated costs are then multiplied by the actual number of first births to teens in those age groups as recorded in vital statistics. All out-year costs are then discounted to future value. These costs are compared to the probable public costs for families in which the mother waited until she was at least 20 before having her first child; the difference is the possible public savings of helping teens delay first births until the age of 20 or later.

Burt and Levy (1987) used these cost analyses as the basis of a model for assessing cost-benefits and the cost-effectiveness of services to adolescents. The reasoning they followed is similar to that used by Cohen (1995), described below for preventing a youth from following the path of a career criminal. That is, we can calculate, using Burt's cost methods, the public cost of each birth to a teenager, and can also estimate what the public would save if that birth were delayed until the teen was 20 or older. If the cost of the program to help each teen avoid an early birth is less than, or even virtually equal to, the cost to society of having that birth occur, then the program is cost-beneficial. If one program can accomplish the same outcomes for teens with less money, then that program is more cost-effective.

Educational Completion

In the United States, roughly one-fourth of youth do not complete high school; in the poorest and most disorganized sections of the country's biggest cities, this statistic can easily double. Catterall (1987) and Chaplin and Lerman (1996) present recent efforts to calculate the social costs of failure to complete at least a high school education, including private costs (those borne by the individual dropping out) and public costs (borne by the society as a whole). The primary component of private costs is foregone earnings as a consequence of being unable to get better-paying jobs and make major advancements in a career. Earnings differences are apparent early, and increase with age. Other private costs are easier to name than to place a dollar value on, and include such things as: more difficulty managing personal finances, less savings and therefore greater vulnerability to crisis; possible lower self-esteem, greater tendency to depression; more exposure to crime and violence (as a consequence of lower incomes and therefore restricted residential choices); over-representation among substance abusers and homeless people known to service providers; greater likelihood of criminal involvement.

Public costs include lost tax revenues (on the foregone earnings), welfare and unemployment services, and possible criminal justice and health costs for the increased probability that high school dropouts will participate in criminal activities and will make greater demands on health services (either for pregnancy and childbearing, addictions, mental health needs, greater propensity toward using or being the victim of violence, or other purposes).

Earnings losses have proved to be the most readily calculated element of dropout costs. Most recent estimates (Chaplin and Lerman 1996) suggest that dropping out of high school entails earnings losses of between $90,000 and $600,000 during the course of a lifetime, with associated lost tax revenues (public costs) of between $30,000 and $200,000. Chaplin and Lerman conclude that "educated guesses suggest that additional losses to society could be equal to or even more than the earnings losses experienced by the individual" (1996, P.1).

Juveniles Likely to Become Career Criminals

Cohen (1995) estimates the social costs associated with a typical career criminal, a typical drug abuser, and a typical high school dropout. Rather than asking the difficult-to-answer question of how many career criminals, drug abusers or dropouts a program has prevented in order to justify its budget, he asks how much we are currently paying in public (government) dollars as a consequence of failing to prevent youth from taking these paths. The answers he develops suggests that if a program could prevent even one youth from becoming a career criminal it would pay for itself many times over. For career criminals, Cohen looks at:

• Victim costs including tangible ones (lost productivity, medical expenses), intangible ones (pain, suffering, and reduced quality of life), and probability of death;
  
  
• Criminal justice costs, including police, investigative, court, and imprisonment costs;
  
  
• Foregone earnings of the youth while incarcerated.
  
  

He makes these calculations for each type of crime, calculates the probable number and mix of crimes perpetrated by the average career criminal, adds up all of these over the course of a criminal career, discounts the total for present monetary value, and arrives at a sum of $1.0 to $1.3 million per career criminal. Similar calculations (with somewhat different components of cost) for chronic drug abusers and high school dropouts produce estimates of $333,000 to $809,000 for the former and $291,000 to $466,000 for the latter. Eliminating duplication (many career criminals are also high school dropouts and heavy drug users), he arrives at an overall estimate of the "monetary value of saving a high risk youth" of $1.5 to $2.0 million. It should be apparent to the reader that even if these estimates are ten times too high, they could still justify a great deal more assistance to high risk youth than currently is being offered.

Why Use a Holistic Approach?

In deciding what to do to assist youth, past efforts have focused largely on youth who have already exhibited behaviors considered undesirable by society (e.g., dropping out of school, engaging in illegal activities, engaging in violent activities, having babies without being able to support them, abusing drugs and alcohol). Programs addressing these issues typically work with youth to try to stop them from continuing these behaviors, and to reduce the consequences of these behaviors. Thus they are engaging in either secondary prevention or tertiary attention (Barker and Fuentes 1995).

There is ample evidence, both from the United States and internationally, that programs of this type cannot and do not bring about a significant reduction in negative behaviors for the youth population as a whole (Barker and Fuentes 1995; Dryfoos 1990; Resnick et al. 1992). The major criticisms of, or explanations for the failure of, most programs of this type are that: (1) they are single focus (they focus only on a single problem, not on the youth's whole life circumstances, and the single problem cannot be solved without "solving" the youth's life circumstances); (2) they are too short-term; (3) they start too late (after the problems are already apparent and the behavioral patterns are set); (4) they focus only on preventing negative behavior, not on promoting positive behaviors. Dryfoos (1990) summarizes the characteristics of single-focus programs with the highest records of success; her findings are reproduced in Table 1.

The evidence is very strong that the successful programs, even though nominally single-focus, in reality treat children and youth (and their families) holistically, start young, offer many enrichment, growth and development activities, and stick with youth for a long period of time.

Recent research results support the importance of an integrated approach to helping youth, both in terms of the issues addressed in the youths' lives and in terms of the structure of services and activities available to them through the intervention program. In New Jersey, a high school that put in place an integrated services program described as "convenient, sensitive, non-stigmatizing, and holistic." The program offers services such as individual and family counseling, primary and preventive health services, drug and alcohol abuse counseling, crisis intervention, employment counseling, training, and placement, summer and part-time job development, recreational activities, and referrals to health and social services. In the year before the program started, students in the school experienced 20 births; in the first year of program operations this number dropped to 13, and to a remarkable 1 in the second year of program operation. Dropouts were reduced from 73 to 24 (of whom 10 received a GED³), and suspensions were reduced from 322 to 78 (Knowlton and Tetelman, 1994).

The Children at Risk Program is a demonstration program operating in six U.S. cities. It offers integrated services and enrichment activities to youth already in trouble with the law at intake (ages 11-13), and residing in the poorest and most crime—ridden neighborhoods of the cities in which the program functions—the program deliberately targets "the worst kids in the worst neighborhoods." First year findings reveal that the program is already making a significant difference (Harrell, 1995a). Compared to a control group (youth were randomly assigned to treatment and control groups in each city), program participants had significantly fewer contacts with police and contacts with the courts, and were significantly more likely to be promoted to the next grade. In addition, trends in school attendance were in the right direction but not significant (reduced absenteeism). Further, community policing aspects of the same projects indicate that the results hold for the entire neighborhood catchment areas of the projects, in comparison to matched neighborhoods in the same cities that had neither the community policing or the Children at Risk Program. Not only were juvenile arrests significantly lower in the demonstration than in the comparison neighborhoods, but so were all criminal offenses, both serious and less serious.

A final U.S. success story is the Quantum Opportunities Program (Hahn 1994; 1995). The evaluation for this program also involved random assignment. The program offered integrated services, enrichment activities, and a community service and development focus. Participants, who were recruited in 9th grade (at about age 14) and followed for four years, received tutoring and other educationally oriented assistance, mentoring (association with compassionate, caring, competent adults), and financial incentives, and also were involved in numerous community-oriented activities, services, and developmental projects. Over the four years during which the study followed the participants, youth participating in the program experienced significantly better outcomes than those in the control group. The results were consistently in the right direction; however, as a warning to those who might hope for "instant," or at least "quick," impact, the differences did not reach the level of statistical significance until the third or fourth year of program participation (some results reached significance sooner than others). These findings suggest that it is important for both program and researchers to be in the picture for an extended period of time, and for the evaluation design to include indicators of early signs of progress, as well as indicators of achievement on the ultimate outcomes toward which the program is targeted.

These results indicate that integrated services addressing the complex needs of youth in a compassionate way with a positive outlook can make a significant difference on important youth outcomes that are likely to affect the future life chances and productivity of youth. The importance of an integrated, holistic approach has now reached the point of acceptance at the level of national government agencies in the United States. For instance, the Office of Juvenile Justice Programs (OJJP) (1995a,b) in the United States Department of Justice has recently issued two reports that accept and strongly promote the holistic approach of involving many other service sectors, families, and communities. OJJDP recognizes that its strictly "justice" goals will not be reached without broader-based interventions, both for primary prevention and for interventions with criminally-involved youth. OJJDP goes so far as to include adequate prenatal care and parenting education as effective strategies for preventing adolescent criminal behavior, both because they can help prevent disabilities that may leave a youth few positive alternatives and because they prevent child abuse that can perpetuate another violent generation. Other U.S. agencies such as the Department of Health and Human Services also take this view.

The International Perspective

A recent discussion paper prepared for the WHO/UNFPA/UNICEF Study Group on Programming for Adolescent Health (1995) strongly supports the holistic approach to youth services, and gives a number of examples from around the world of youth-serving programs that take this approach. One of these, Proyecto Alternativas in Honduras, serves working children in the informal sector and their families, as well as street children without families. It combines educational and social services, community-based health care, food supplementation, and basic and health education including a focus on the issues of sexuality, reproductive health and substance abuse, counseling, and significant levels of youth empowerment and control over the project's planning and decision-making. Barker and Fuentes (1995) describe numerous youth-serving programs with integrative components; most operate within Latin America and the Caribbean. One example is Servol in Trinidad and Tobago. Primarily an employment and training project that now offers services throughout the country, Servol found that it had to incorporate a life skills component in order to help youth develop the personal skills they need to successfully function in the society and the workplace. These include components from many service sectors, such as: self-awareness; parenting; nutrition, health and sex education; drug abuse prevention; sports and recreation; basic literacy and social studies; community service; becoming a micro-entrepreneur; and, obtaining immediate employment. Youth must complete the life skills component before receiving formal job training. Gender equality in the program and in future work is an explicit goal of Servol. Unfortunately, basic evaluative assessments of these programs have not been done, although they are perceived to be successful at achieving their goals.

---

Recommendations for Shaping and Targeting Investments in Adolescents

In the pursuit of a more holistic approach, it is not necessary, and may not even be desirable, for a particular program to try to do everything itself.

As should be obvious from the previous section, it is a strong recommendation of this paper that programs structure themselves so they can serve youth in a holistic manner. The keys to doing so are to take a long-range approach, to offer positive developmental opportunities to youth appropriate to their age and experience, to pay attention to family and neighborhood issues, and to recognize that healthy behaviors (and unhealthy behaviors) happen in a context which must be addressed if programs are to be successful in changing the lives of youth. This may be a new way of thinking for many readers of this paper, but the evidence is coming in from many directions to suggest that the payoffs are worth the effort. In thinking about the value of the holistic approach, one needs to think about the reasons why youth may not benefit from your program, and determine to help them overcome these barriers. The barriers may be those of knowledge and access, but they may also be those of motivation, of seeing no realistic future that is worth the effort to take care of themselves, of other pressing priorities, of no resources to follow the advice of the program on health issues, of family or peer interference, or of other neighborhood influences. If 90 percent of the youth in your catchment area experience one or more of these barriers to the extent that they are prevented from using your program, it may be wise to develop ways to reach this 90 percent, rather than working exclusively with the 10 percent who are able to find their way to your door.

In the pursuit of a more holistic approach, it is not necessary, and may not even be desirable, for a particular program to try to do everything itself. More and more, programs addressing different needs and providing different skills and knowledge are forming networks that enable them to meet the multiple needs of youth in a flexible and efficient manner. This may involve co-locating services from different agencies in an attractive and easily accessible youth center, or it may involve developing a network of services that are very familiar with each other's work and make a lot of referrals among network agencies. It certainly involves listening to the needs and issues of youth, the needs and issues of service providers, and creating new and flexible structures that are responsive to both sets of needs.

Challenges in Establishing Strong Comprehensive Programs⁴

Establishing Priorities and Getting Commitment

As the first concrete step in the planning process, the services, agencies, funders, potential youth participants, and other partners involved should work toward agreement on a common set of goals and objectives (Center for the Future of Children 1992). To the extent possible, long-term commitment to the effort should be built in from the planning stage. One effective method for encouraging long-term commitment is through an independent inter-agency advisory group with a revolving chair, to help minimize turf battles and forge a common purpose for the variety of service integration partners. Another method involves diversion of a portion of each partner's funds to support the effort, so each partner has an important stake in assuring success of the integrated approach.

The program's goals should be based on a local community needs assessment and an assessment of services already available, whether formal or informal. If the full range of stakeholders is included in the planning process, including a critical role for youth themselves, knowledge of service needs and adequacy of existing services should be included. Efforts should be made to solicit input and build support from as many of the partners as possible. Outside consultants can also be brought in to share their expertise (Corriea, 1992).

When thinking about who should be involved in this stage of developing commitment, it is important that all levels of critical agencies be at the table. Agency directors are the ones who can commit funds, but section supervisors will be the ones with the day-to-day management tasks, and front-line workers are the ones who know all the problems that the effort will encounter. Each has an important role to play in defining goals and objectives. The voices of youth, families, and community actors are also critically important to include, as they can help shape the program so that people will want to and be able to use it.

Getting Started and Initial Implementation

Identifying the Target Population

Identifying the Target Population. Who should the new comprehensive, holistic services be designed to help? Unless the target population is clear, it will not be obvious what services and other activities should be incorporated into the effort. Whether services should be offered to all youth and families in the community, or only to those considered at highest risk, is an important policy question for local partners to address (Levy and Shepardson 1992). Once a youth or family enters the program, agencies should have sufficient knowledge of services available, inter-agency cooperation, and flexibility to ensure that all of their service needs are identified and addressed. Some authorities maintain that services should be concentrated on those who are most at risk; others argue that this approach would stigmatize program participants, and that all children could benefit from enrichment efforts (Dryfoos 1990).

The U.S. Office of Technology Assessment (1991) concluded that adolescents who are not currently being served by the myriad of prevention and treatment programs are those "with, or at risk of multiple problems, who almost inevitably face gaps among service systems" (p. I-30). Adolescents most likely to encounter service gaps are those with substance abuse and mental health problems, adolescents adjudicated as delinquent but who probably have multiple health problems, homeless adolescents, and adolescents failing or misbehaving in school who are also likely to become pregnant, delinquent, and/or drop out of school.

If a program targets 10- to 15-year-olds and their families, a different array of activities and services are likely to be needed than if an older adolescent population were the target. For the younger group, prevention activities involving recreation, community service, self-esteem and competence building, compensatory educational efforts and similar activities will be primary, with treatment services on reserve and accessible if needed. Older youth may need a stronger mix of treatment services to help them stop participating in problem behaviors, as well as the supportive developmental services offered to younger teens.

Identifying the Services to be Offered

A comprehensive approach involves a child- and family-centered orientation approach in which the range of each family's service needs are identified and services are planned and delivered to address their unique situation. This contrasts with a problem-centered approach, in which an agency addresses only the specific problems it is prepared to handle itself. A comprehensive approach requires considerable variety in the breadth and depth of services available and flexibility in service delivery. It is always important to remember, however, that "comprehensive" and "integrated" are not identical. The point of developing a service structure is not to assemble the largest number of services, but to help youth and their families. Successful youth service programs are marked by their common emphasis on client empowerment rather than on narrowly defined "services" from public agencies (Pittman and Cahill 1992). However, the system must have the ability to get clients the services they need from all agencies in the community.

The type of services to be offered, including outreach, public education, primary and/or secondary prevention, intervention, and advocacy, needs to be decided on the basis of local needs and resources. With young adolescents, primary and secondary prevention is likely to be a major focus.

The breadth of services is another issue. In one view, a minimum of two specific types of services in each of the three broad categories of education, health, and social services should be offered for the program to be considered truly comprehensive (Morrill and Gerry 1990). Others argue that basic life skills such as critical thinking, problem-solving, and decision-making, social skills such as constructive assertiveness, and the use of social support systems should be the program's focus (Hechinger 1992).

The intensity of services should also be considered. The service programs should be flexible enough to respond to clients who may require more frequent services or services that address the relevant issues in more detail.

For the target population of young adolescents, there is also some question about the best way to provide comprehensive services. The more a program emphasizes prevention, the more it may focus on developing self-esteem and positive life skills, resisting peer pressure to participate in risky behaviors, and fostering a belief that youth can have a positive and productive future as an adult. Programs may promote these goals through emotionally supportive role-modeling from mentors or big brothers/sisters. A comprehensive program in this context would assure that the mentor has access to someone in a case management role when it becomes apparent that a youth needs a particular type of help. In contrast, a program that involves heavy up-front assessment and case management may be more appropriate for the small proportion of 10- to 15-year-olds who need massive early intervention, or for older adolescents who are in trouble already.

Mechanisms for Service Delivery

The way in which services are coordinated is important. Clients may have a service agency contact with whom they maintain an ongoing, supportive relationship. When this contact person functions more as a mentor, counselor, or group worker than as a case manager, this individual needs access to someone who can arrange needed services and follow up on referrals.

Case management—a key issue—is essentially a method of placing responsibility for service planning, coordination of service delivery, and follow up on an individual or inter-agency team. The case manager or team works with youth and their families to determine service needs, provide inter-agency linkages, and monitor service delivery and outcomes (Melaville and Blank 1991). Effective case management requires relatively smaller caseloads as the needs of clients increase. Intensity of services offered should be determined at least in part by the youth and family's ability and motivation to work with the system. The procedures established should be flexible enough to respond to each youth and family's unique circumstances.

Service Location

Coordinated, comprehensive services can be delivered through school-based or school-linked sites, in community sites such as churches or community centers, through mobile arrangements, and/or by home visits. There is no universally applicable program model. In all likelihood the location of the effort will depend on which agency or organization has an interested, committed, and dynamic person willing to take the lead in developing and running the program. Another important factor is the site's acceptance within the community. Occasionally a local agency may get involved because some funding source has invited its participation. Such invitations are most likely to be accepted when there is local leadership to carry the program.

Services are typically based in either school or community sites. School-based programs have the potential to reach large numbers of youth, and have a well-established organizational structure and niche in the community, but may not be as accessible to families or to youth and families who are alienated from the educational system, such as high risk dropout youth. They may also further stress an overburdened educational system (Chaskin and Richman 1992), may be restricted as to which services they can provide (e.g., some are unable to provide family planning services), and may be constrained by rigid organizational rules. Community-based programs may avoid these problems but face issues of access for youth and families and high-crime and gang-infested neighborhoods.

Debates also occur about the appropriate balance of services between on-site and off-site locations. Some programs aspire to on-site "one-stop-shopping," while others function as a link between clients and a very broad spectrum of services"none of which is offered on site. The debate about service concentration usually involves the relative benefits of ease of access versus learning to negotiate the systems oneself. Most programs fall somewhere between these two extremes. A community just beginning to develop a coordinated approach should consider this issue.

Administrative Factors

To be a credible model of coordinated, comprehensive services, the programs involved should have institutionalized linkages that establish the mechanisms for sharing resources. These mechanisms may include co-locating in a single facility; sharing staff, financial resources, and/or information; and agreeing to provide services to referred people. An agency that provides needs assessments, service referrals, and referral follow-ups must be able to give referral agencies the information it has about a client's needs. Many agencies have confidentiality policies that prohibit the disclosure of client information between service agencies, and sometimes even within different divisions of a single agency. For the coordination effort to work, agencies must find ways to adjust these confidentiality policies and still protect sensitive information about clients. Gaining the informed consent of clients to share information with agency personnel who will be providing the referral service is one approach that has worked in some places. But even this may require formal legal or rule changes.

Staffing Issues

It is important that staff be recruited and trained very carefully, whether they are paid or unpaid (Primm Brown 1992). Staff should be selected on the basis of their ability to establish trusting, respectful relationships with youth and families, their ability to span professional boundaries and specializations to address clients' needs, and their ability to work with the system, whatever their type or level of professional training (Sonenstein et al. 1991).

Diversity issues must also be considered in staffing programs (Corriea 1992). If at all possible, staff should reflect the racial, ethnic, age, and gender make-up of the program's clientele. At an absolute minimum staff should have a demonstrated sensitivity to issues of racial, ethnic, and gender diversity, preferably through earlier work experience with populations similar to those expected to use the program.

Staff support for the effort and willingness to adopt new roles are crucial at all levels. Strong positive leadership is usually critical; neutrality is not good enough to shepherd a new program to successful implementation.

Staff at all levels should be trained to work effectively to serve the goals of the new approach. Training should be sensitive to the concerns of staff experienced in non-integrated service settings—concerns such as "turf" issues, professional orientations and jargons, and youth and family issues that staff may feel unprepared to deal with.

Funding Issues

For a coordinated comprehensive approach to work best, funding should be flexible. Funding sources should be redesigned to blend together funds from multiple sources that historically have had rigid categorical boundaries, so programs will be able to provide adequate and coherent funding for services that address multiple areas of need (Kirst, 1991). One promising approach to increasing service coordination among already functioning programs is to use limited new funding to support core coordinating functions. This effort could be matched by diverting some existing funds to support additional efforts and using other existing funds to support regular service delivery.

Maintaining Program Integrity

A long-term issue is whether any changes created by the new effort in the component agencies' functioning and interrelationships become institutionalized and take on a life of their own. Kusserow (1991), summarizing twenty years of attempts to create coordinated service structures, notes that "... efforts have been instrumental in making human services more accessible to clients and more responsive to their needs. Over the long term, however, efforts appear to have had little institutional impact on a highly fragmented human services system." His list of major barriers to system change are:

• The size and complexity of the human services system;
  
  
• Professionalization, specialization, and bureaucratization;
  
  
• Limited influence of those trying to create change;
  
  
• Weak constituency for service integration;
  
  
• Funding limitations; and
  
  
• Insufficient knowledge.
  
  

It is very important that efforts for change rest on more than seed funding and strong personalities or leadership. Such factors are likely to be transitory. A program depending on these factors is likely to collapse when the funding expires and the individuals depart. Pooling at least a portion of each agency's core funding to support coordinated activities is a systemic change that can be crucial in assuring the survival of the coordinated service network. This practice may assure adequate resources to continue the coordinated approach after start-up funding expires. It may also solidify the commitment of participating agencies by their very tangible stake in the new service structure (Melaville and Blank 1991).

Where long-range funding is inadequate to sustain the coordinated, holistic approach, the availability of evaluation data documenting the innovative processes and beneficial outcomes resulting from the use of the new approach can be instrumental in securing continuation funding (Melaville and Blank 1991). Policymakers and (potential) funders can make better-informed decisions about how to allocate limited resources when information is available to document implementation procedures, service costs, and cost savings. Even more desirable is information showing the impact of the new approach on program participants, component agencies, and the social service system.

Achieving Significant Youth Involvement

Achieving Significant Youth Involvement. As has been reiterated throughout this paper, the involvement of youth is critical at every step along the way. Youth should be involved in helping to identify needs and set goals, in helping to design service and activity structures that will be attractive and accessible to youth, in helping staff understand what youth need and how to relate to youth, in identifying the positive activities and supports that will attract youth to the center and keep them coming, in learning more about the community and family issues that are barriers for youth and in creating the environment in which youth can become positive contributors to their community. Some service systems have set up councils of youth advisors; others have turned over some significant amount of program control to youth in a variety of ways. Many have involved youth as helpers of other youth, either as tutors, mentors, peer counselors, information disseminators, organizers, or role models. When youth are involved in program design and management in these ways, their presence and the respect they are given help draw other youth to the program, and empower all youth.

Getting Evaluation and Other Vital Information

There are a number of reasons why it may be important to collect information about or to evaluate programs for at-risk youth. The first of these is that such information can help you learn how to start programs and implement them successfully. The second is that through the process aspects of evaluation activities you can learn how existing programs may be improved. Third, you can document potential program benefits to youth and to society. Finally, data collection and evaluation activities can help you justify program costs and show that programs avert future expenses through their prevention and intervention activities.

In the remainder of this paper, I only have space to refer briefly to the many issues and options you have when you decide to gather information about program performance. These issues are much more thoroughly covered by Harrell (1995b) in a paper prepared for the World Bank on evaluation of youth-serving programs in Latin America.⁵

Most people usually think about three phases or levels of evaluation:

• Implementation: getting started, reaching a steady state, making changes;
  
  
• Process: daily activities, service delivery, participation in a mature, stable program;
  
  
• Outcome, impact, effects: reductions in individual and community risk behaviors and negative outcomes; increases in positive behaviors.
  
  

To these we add three other types of data collection that may be even more useful in the short run:

• Performance monitoring: routinely collecting program data on many of the same issues covered by a formal evaluation, for the purpose of obtaining feedback about program performance;
  
  
• National level-data collection: routine collection of basic data reflecting the status and achievements of youth (e.g., school completions, employment, living situation, basic health and vital statistics data);
  
  
• Costs: of running the program, and of achieving outcomes.
  
  

Choosing an Evaluation Design

Why Do a Case Study or an Ethnograpy?

Case studies or ethnographies are the "softest" types of evaluation activity. They generally produce qualitative data about program operations and program clients. If well done, however, they are often very informative in a number of ways. An ethnography can tell you what potential participants need and what they want. It can tell you who needs to be involved in the program (e.g., youth only, families, etc.). It can tell you where to put the program so it will reach desired participants. And it can tell you how the program relates to the rest of the community. It is often the first method people use when they are thinking about program changes. A case study can help you understand how the program develops and how it actually operates. It is a very good methodology for gaining an understanding of the barriers to success and strategies for overcoming them. And it can provide feedback for improving the program or replicating it somewhere else. Many of the techniques of a case study are the same as those used for implementation and process evaluation, and have the same benefits.

Why Do Performance Monitoring?

Performance monitoring is systematic, routine, repeated data collection of information about services and program outputs. For services, one often monitors the volume of services (how many?) and the efficiency of their delivery (how quickly, with how much client waiting time, with how much staff down-time, etc.?). For outputs, one often monitors the number of clients who complete the recommended array of services, the time it takes them to do so, what sequence they follow in doing so, and their satisfaction with services.

Performance monitoring can help you describe your program's performance quantitatively, with respect to the number of each type of service given, the number of participants per month, or the number of days per month that each participant uses the program. It can provide feedback on success by comparing performance to numerically stated goals, and can monitor service quality through participant surveys. Finally, you can use performance monitoring information together with information about costs and resources consumed to assess efficiency, cost-effectiveness, and productivity.

Why Do Impact Evaluations?

An impact design helps you document what the program has accomplished, see whether the program is meeting its ultimate goals of helping youth even after they leave the program, convince funders that the program is worth supporting, and encourage program replication. An impact design can be experimental (using random assignment and true control groups), quasi-experimental (without random assignment, but with one or more comparison groups to protect against a variety of alternative explanations of results), or non-experimental (without either random assignment or formal comparison groups, but—hopefully—with other means of checking for bias, such as measures taken over multiple time periods before and after program changes have been made that are expected to impact results).

There are various issues that must be resolved in the process of selecting an impact evaluation design. Recognizing that most programs and communities are not in a position to support the "perfect" evaluation design, nevertheless there are many useful and informative designs available. One of them will probably fit the circumstances of your program or community. The type of design you end up with will be contingent on:

• Whether you can do random assignment of youth, of classes, or of any other unit;
  
  
• Whether you can find a reasonable and convincing comparison group;
  
  
• Whether you can get—before and after—measurements;
  
  
• Whether you can take measurements on the same cohort of youth frequently; and
  
  
• Whether you can follow the participants (and control/comparison group members) long enough to observe your ultimate outcomes.
  
  

Finally, once a design has been selected, there are other issues which the evaluator must consider, and about which decisions must be made. These include:

• Identifying the clients to be included, and defining the unit of analysis. Not every youth who walks through your doors receives the same level of service, or remains attached to the program for the same length of time. Which ones do you want to count as "clients"? Perhaps you have several types of clients, such as those who participate primarily in sports and recreational activities, and those who receive many more remedial services. You may want to treat them separately in the evaluation.
  
  
• Assessing differing levels of client risk at program entry. Some clients are very troubled, others are not. If your program deals primarily with multi-problem (high risk) clients, whereas another program deals primarily with clients who have few problems and thus absorb fewer services, your program will end up looking more like a failure given the same amount of resources committed to it, unless you can show that your clients were harder to serve successfully.
  
  
• Deciding what you may want to know about system change, as well as individual change. If you want to know that referrals got faster, that new types of services or methods of service delivery were put in place and successfully used, you need to set up ways to document this.
  
  
• Documenting service delivery. Did some clients achieve successful outcomes with relatively few services? Did other clients fail even though they got a lot of services? Does the amount and mix of services make a difference for client outcomes? You won't know unless you collect information on services received by each client, and relate the services to client outcomes.
  
  
• Deciding who should conduct the evaluation. Usually it is better to have an outsider do it, both for the credibility of the evaluation and because program staff rarely have time to divert to evaluation activities. Program staff must be deeply involved, however, since most of the day-to-day recordkeeping will fall to them.
  
  
• Deciding what outcomes to measure. These should relate to your program=s goals and objectives, but be realistic given the time frame in which you are working. Selecting goals to measure that take 5-10 years to achieve, when you only have two years to do the evaluation, is a recipe for failure.
  
  
• Finding a comparison/control group. If you can, great. If you can't, you will have to construct your design to compensate.
  
  
• Reducing attrition at follow-up. If you cannot find half of your failures when you go to collect your ultimate outcome data, but you can find all but 10 percent of your successes, your program is going to look very good. If the figures are reversed (because your failures are still coming to the program but your successes have moved on), your program is going to look very bad. It is extremely important to get everyone into the follow-up data collection, or at least to eliminate bias in the people you can reach.
  
  

Why Do Cost Analyses?

The usual type of cost analysis summarizes all program costs and describes the major cost centers of the program. If one is interested in per-client costs, one can then divide total program costs by the number of clients served. It may also be important to include non-monetary costs, such as volunteer time, donated building space and materials, and other such factors, because other communities interested in duplicating your program will need to know what it really took to make the program work. If half of your resources were non-monetary, they need to know it, or they will fail.

If you know what your program costs, per client per year, you can compare this with the analyses you have done on the savings from rescuing a high risk youth from a life of crime, drug addiction or incomplete education, and justify your program against those undoubtedly much larger costs. This would be a particularly convincing argument to make if your evaluation data were able to show that through the efforts of your program, some significant number of your clients did, in fact, avoid negative outcomes and go on to be productive members of adult society.

---

Summary and Conclusions

Youth behavior does not occur in a vacuum, and cannot be fixed in a vacuum. It occurs in the context of the youth's family and family dynamics, peer group, and neighborhood and social opportunities.

This paper has argued that it is important for developing countries to invest in adolescents, both to prevent the negative consequences of potential risky behavior and to promote healthy growth, modern skills, and the capacity to participate in the society of the future. It has also argued that the most effective way to make such investments is to think and act holistically and from a preventive and developmental outlook. The following is a summary of the major points made throughout the paper:

The Need for Future Health and Productivity

• Adolescents aged 10-19 are a very significant proportion of the population of Latin American and Caribbean countries, averaging 21.7 percent.
  
  
• The future economic development of Latin American and Caribbean countries depends on having increasing proportions of the population who are reasonably well-educated, healthy, and economically productive.
  
  
• Many Latin American and Caribbean youth do not complete enough education to equip them for productive labor in the formal modern sector—this situation is worst among poor urban youth and among rural youth of all income levels.
  
  
• For many young adolescents, employment may be necessary to help family finances, but it interferes significantly with educational attainment. Conversely, many older adolescents cannot find adequate work and are neither in school nor employed.
  
  
• Teenage sexuality, pregnancy, and childbearing present major health issues; the health risks to young women from illegal abortions are also great.
  
  
• The consequences of violence—from neighborhood violence and gangs, from household members (child abuse and domestic violence), and from war—are also severe in many countries of the region.
  
  

The Rationale for Investing in Youth

• There are substantial costs to governments and to individuals for every failure of youth to reach adulthood alive and healthy, with an adequate education, and without children they cannot care for.
  
  
• These costs are almost always greater—sometimes vastly greater—than the costs of programs to help youth achieve these goals.
  
  
• Some of the costs have been calculated for the United States, and are of surprising magnitude:
  
  

–	Each year's group of youth who do not finish high school will, over their lifetime, cost the nation $260 billion in lost earnings and foregone taxes.
–	Each year, the U.S. spends roughly $20 billion in payments for income maintenance, health care, and nutrition to support families begun by teenagers.
–	The cost of one career criminal has been calculated as between $1.0 and $1.3 million; similar calculations for one chronic substance abuser are $333,000 to $809,000.

The Framework for Youth-Serving Activities

• Youth behavior does not occur in a vacuum, and cannot be fixed in a vacuum. It occurs in the context of the youth's family and family dynamics, peer group, and neighborhood and social opportunities. The more desperate the context, the more supports youth will need to survive and prosper.
  
  
• Programs that focus on fixing single problems, however serious–e.g., school dropout, drug abuse, teenage childbearing, criminal behavior, unemployment–cannot and do not change the lives of youth. They treat only the symptom, not the underlying problems.
  
  
• Programs need to: (1) start early; (2) work with youth for years; and (3) holistically and comprehensively address the needs and aspirations of youth, including those of their families, peers, and neighborhoods.
  
  
• In addition to helping with concrete problems, programs need to promote positive behaviors, offer activities and opportunities that give youth reasons to expect a decent future, and promote the capacity for participation and self-determination.
  
  
• The opportunity to develop stable relationships with competent and caring adults is an especially critical component of successful programs.
  
  
• Example: Servol in Trinidad and Tobago is an employment and training project offering services throughout the country. Servol found that it had to add a life skills component to help youth develop the personal skills they need to function successfully in the society and the workplace, including: self-awareness; parenting; nutrition, health and sex education; drug abuse prevention; sports and recreation; basic literacy and social studies; community service; becoming a micro-entrepreneur; obtaining immediate employment. Youth must complete the life skills component before receiving formal job training.
  
  

Steps in Establishing Strong Comprehensive Programs

• Establishing priorities and getting commitment
  
  
• Identifying target population(s) and involving youth in planning
  
  
• Deciding which services and activities to offer, where they will be located, and how they will be coordinated
  
  
• Handling administrative, staffing, and funding issues
  
  
• Maintaining program integrity
  
  

Getting Evaluation and Other Vital Information

• Information can help you learn how to start programs, implement them successfully, and replicate them in new sites
  
  
• The process components of evaluations can identify how existing programs may be improved
  
  
• The program can document its benefits to youth and to society
  
  
• Data collection and evaluation activities can help justify program costs and show that programs avert future expenses through their prevention and intervention activities.
  
  

---

References

Barker, G. and Fuentes, M. 1995. "Review and Analysis of International Experience with programs Targeted on at-Risk Youth". Washington, D.C. Paper prepared for the World Bank, Human Resources Division, Country Department III, Latin America and the Caribbean Region.

Blos, P. 1962. On Adolescence. New York: Free Press of Glencoe.

Burt, M.R., Resnick, G. and Matheson, N. 1992. "Comprehensive Service Integration Programs for at-Risk Youth". Final Report. Washington, DC: Urban Institute.

Burt, M.R. 1986. "Estimating the Public Costs of Teenage Childbearing". Family Planning Perspectives, 18(5), 221-226.

Burt, M.R. 1985. Teenage Pregnancy: How Much Does It Cost? Washington, DC: Center for Policy Options.

Burt, M.R. and Levy, F. 1987. "Estimates of Public Costs for Teenage Childbearing: A Review of Recent Studies and Estimates of 1985 Public Costs". In C. Hayes (Ed.), Risking the Future: Adolescent Sexuality, Pregnancy and Childbearing, Vol. II. Washington, DC: National Academy Press.

Campbell, A. 1987. "Self-reported delinquency and home life: Evidence from a sample of British girls". Journal of Youth and Adolescence, 16, 167-177.

Carnegie Council on Adolescent Development. 1983. Turning Points: Preparing American Youth for the 21st Century. Washington, DC: Carnegie.

Catalano, R.F. and Hawkins, J.D. 1995. Risk-focused Prevention: Using the Social Development Strategy. Seattle, WA: Developmental Research and Programs, Inc.

Catterall, J. 1987. "On the social costs of dropping out of school." The High School Journal, 71, 19-30.

Center for the Future of Children. 1992. Analysis. In R. Behrman (Ed.), The Future of Children: School-linked Services, (1), 6-18.

CEPAL 1994. Panorama Social de Am&#eacute;rica Latina. Santiago de Chile: Naciones Unidas.

Chaskin, R.J., and Richman, H.A. 1992. "Concerns About School-linked Services: Institution-based Versus Community-based Models". In R. Behrman (Ed.), The Future of Children: School-linked Services, (1), 107-117.

Corriea, P. 1992. National Resource Center for Youth Services, Tulsa, Oklahoma. Phone interview with author, May 20, 1992.

Chaplin, D. and Lerman, R. 1996. Dropping out: Public and private costs. Washington, DC: Urban Institute.

Cohen, M.A. 1995. The monetary value of saving a high risk youth. Washington, DC: The Urban Institute.

Cowen, E.L. and Work, W.C. 1988. "Resilient children, psychological wellness, and primary prevention." American Journal of Counseling Psychology, 16, 591-607.

Dryfoos, Joy. 1990. Adolescents at Risk. New York: Oxford University Press.

Felsman, J.D. and Vaillant, G. 1987. "Resilient Children and Adults." Advances 4 (4), 45 - 61.

Felner, R.D., Aber, M.S., Primavera, J. and Cauce, A.N. 1985. Adaptation and vulnerability in high-risk adolescents: An examination of environmental mediators. American Journal of Community Psychology, 13 et al, 365-379.

Flores, C.E., Knaul, F.M. and M&#eacute;ndez, R. 1994. "Un an&#aacute;lisis cuantitativo del trabajo infantil y juvenil en Colombia". Desarrolo y sociedad, 34.

Freud, A. 1958. "Adolescence". Psychoanalytic Study of the Child, 16, 225-278.

Garmezy, N. 1987. "Stress, competence, and development: Continuities in the study of schizophrenic adults, children vulnerable to psychopathology, and the search for stress-resistant children". American Journal of Orthopsychiatry, 57, 159-174.

Garmezy, N. 1983. "Stressors in childhood". In Garmezy, N. and Rutter, M. (Eds.), Stress, Coping and Development in Children. New York: McGraw-Hill.

Grossman, F.K., Beinashowitz, J., Anderson, L. Sakurai, M. Finnin, L. and Flaherty, M. 1992. "Risk and resilience in young adolescents". Journal of Youth and Adolescence, 21(5), 529-550.

Hahn, A. 1995. "Quantum Opportunities Program: A Brief on the QOP Pilot Program". Waltham, MA: Brandeis University, Heller School, Center for Human Resources.

Hahn, A. 1994. "Evaluation of the Quantum Opportunities Program (QOP): Did the Program Work?" Waltham, MA: Brandeis University, Heller School, Center for Human Resources.

Hamburg, B.A. and Takanishi, M. 1989. "Preparing for life: The critical transition of adolescence". American Psychologist, 44, 825-827.

Harrell, A.V. 1995a. Impact of the Children at Risk Program: Preliminary Findings from the First Year. Washington, DC: Urban Institute.

Harrell, A.V. with Burt, M.R., Hatry, H., Rossman, S., Roth, J, and Sabol, W. 1995b. "Evaluating programs for vulnerable children and youth." Washington, DC. Paper prepared for the World Bank, Human Resources Division, Country Department III, Latin America and the Caribbean Region.

Hauser, S.T., Vieyra, M.A.B., Jacobson, A.M. and Wertlieb, D. 1985. Vulnerability and resilience in adolescence: views form the family. Journal of Early Adolescence, 5. 81-100.

Hayes, C. 1987. Risking the Future: Adolescent Sexuality, Pregnancy and Childbearing, Vol. I. Washington, DC: National Academy Press.

Hechinger, F.M. 1992. Fateful Choices: Healthy Youth for the 21st Century. New York: Carnegie Corporation.

Heise, L., Pitanguy, J. and Germain, A. 1994. "Violence against Women: The Hidden Health Burden." Washington, DC: World Bank, Discussion Paper 255.

Jones, E., Forrest, J.D., Goldman, N., Henshaw, S., Lincoln, R., Rosoff, J., Westoff, C., and Wulf, D. 1985. "Teenage pregnancy in developed countries: Determinants and policy implications." Family Planning Perspectives, 17, 53-63.

Kirst, M.W. 1991. "Financing School-Linked Services." Palo Alto, CA: Stanford University, unpublished manuscript.

Knaul, F.M. 1995. "Young workers, street life and gender: The effect of education and work experience on earnings in Colombia." Doctoral dissertation, Harvard University, Department of Economics.

Knaul, F.M. and Flores C.E., 1996. "Targeting youth: Empirical evidence, conceptual issues and rationales." Washington, DC. Paper prepared for the World Bank, Human Resources Division, Country Department III, Latin America and the Caribbean Region.

Knowlton, R.B. and Tetelman, E.H. 1994. "Educators Respond to New Jersey's 'One-Stop-Shopping' Program." In R.A. Levin, (Ed.), Greater than the Sum: Professionals in a Comprehensive Services Model. Chicago: University of Illinois at Chicago, Teacher Education Monograph No. 17.

Kusserow, R. 1991. Services Integration: A Twenty-Year Retrospective. Washington, DC: Department of Health and Human Services, Office of the Inspector General.

Kwakman, A.M., Zuiker, F.A.J.M., Schippers, G.M. and de Wuffel, F.J. 1988. "Drinking behavior, drinking attitudes, and attachment relationships of adolescents." Journal of Youth and Adolescence, 17, 247-253.

Levy, J.E., and Shepardson, W. 1992. "Look at current school-linked efforts." In R. Behrman (Ed.), The Future of Children: School-linked Services, 2 (1), 44-55.

Melaville. A.I., and Blank, M.J. 1991. What It Takes: Structuring Interagency Partnerships to Connect Children and Families with Comprehensive Services. Washington, DC: Education and Human Services Consortium.

Morrill, W.A., and Gerry, M.H. 1990. "Integration and Coordination of Services for School-Aged Children." Unpublished manuscript.

Murphy, L.B. and Moriarity, A.E. 1976. Vulnerability, Coping and Growth. New Haven, CT: Yale University Press.

Offer, D. 1987. "The mystery of adolescence." Adolescent Psychiatry, 14, 7-27.

Offer, D., Ostrov, E. and Howard, K.J. 1981. "The mental health professional's concept of the normal adolescent." Archives of General Psychiatry, 38, 149-152.

Office of Juvenile Justice Programming. 1995a. Delinquency Prevention Works: Program Summary. Washington, DC: Department of Justice, Office of Juvenile Justice Programming.

Office of Juvenile Justice Programming. 1995b. Guide for Implementing the Comprehensive Strategy for Serious, Violent, and Chronic Juvenile Offenders. Washington, DC: Department of Justice, Office of Juvenile Justice Programming.

Office of Technology Assessment. 1991. Adolescent Health, Vols. 1, 2, and 3. Washington, DC: U.S. Government Printing Office.

Pittman, K. and Cahill, M. 1992. "Getting Beyond the C's: The Role of Comprehensive Programs, Case Management Services and Coordinated Community Planning in the Development of Supportive Environments for Youth." Academy for Educational Development, Center for Youth. Development and Policy Research, Washington, DC and New York.

Primm Brown, G. 1992. Education and Healthy Development of Children and Youth, Carnegie Corporation. New York, New York. Phone interview with author, May 8, 1992.

OPS. 1990. Las Condiciones de Salud en las Am&#eacute;ricas. Vol I. Washington, DC: OPS, Oficina Sanitaria Panamericana, Oficina Regional de la OMS.

Resnick, G., Burt, M.R., Newmark, L. and Reilly, L. 1992. "Youth at Risk: Definitions, Prevalence, and Approaches to Service Delivery". Washington: DC. Urban Institute.

Robertson, J.F. and Simons, R.L. 1989. Family factors, self-esteem, and adolescent depression. Journal of Marriage and the Family, 51, 125-138.

Rutter, M. 1987. "Psychosocial resilience and protective mechanisms." American Journal of Orthopsychiatry, 57, 316-331.

Rutter, M. 1979. "Maternal deprivation, 1972-1978. New findings, new concepts, new approaches". Child Development, 50, 283-305.

Singh, S. and Wulf, D. 1990. Adolescentes de Hoy, Padres del Ma&#ntilde;ana: Un Perfil de las Am&#eacute;ricas. New York: Alan Guttmacher Institute.

Sonenstein, F.L. Ku, L., Juffras, J. And Cohen, B. 1991. Promising Prevention Programs for Children. Alexandria, VA: United Way of America.

SRI International. 1979. An analysis of government expenditures consequent on teenage childbearing. Menlo Park, CA.

Werner, E.E. 1989. "High-risk children in young adulthood: A longitudinal study from birth to 32 years". American Journal of Orthopsychiatry, 59, 72-81.

Werner, E.E. 1986. "Resilient offspring of alcoholics: A longitudinal study from birth to age 18". Journal of Studies on Alcoholism, 47, 34-40.

Werner, E.E. and Smith, R.S. 1982. Vulnerable but not invincible: A study of Resilient children. New York, McGraw-Hill.

Wertheimer, R. and Moore, K.A. 1982. Teenage Childbearing: Public Sector Costs. Washington, DC: Urban Institute.

WHO/UNFPA/UNICEF Study Group on Programming for Adolescent Health. 1995. Programming for Adolescent Health: Discussion Paper. Saillon, Switzerland, November 28-December 4, 1995.

---

Tables

Table 1. Program elements that are important to successful youth-serving programs summarizing Dryfoos' 1990 reviews of many programs

	Delinquency Prevention	Sub. Abuse Prevention	Pregnancy Prevention	School Drop-out Prevention
Early intervention, at least by 8,9,10
Long time frame, stick with the kids from pre-adolescence through successful transition to adulthood
Intensive personal involvement is possible, strong bonds with staff and staff with enough time to focus on each individual youth
School involvement—either program located in schools, or strongly connected to schools
Broad goals, offer many life options, offer something to give kids reason to expect decent future
Offer multiple services, be able to treat youth and family holistically, be flexible
Change institutions—create linkages,make multi-agency arrangements, integrate services
Maintain program integrity, do not water it down, do not compromise for financial political reasons

---

Notes

1. This section draws heavily on Burt, Resnick and Matheson (1992).

2. Income assistance (Aid to Families with Dependent Children), health care (Medicaid), and nutrition (Food Stamps).

3. Equivalent to a high school diploma.

4. This section draws heavily on Burt, Resnick and Metheson (1992).

5. Also available in a slightly revised form for general human services programs from the Urban Institute as Harrell, Al, Evaluation Strategies for Human Services Programs: "Guide for Policymakers and Providers." Washington, D.C.: Urban Institute, 1996.

---

Usage and reprints: Most publications may be downloaded free of charge from the web site and may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact [email protected].

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.

Disclaimer: The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute.