• Findings of Change and Stability
• Regrettable Limitations of Coming of Age in the 1990s
• Concluding Remarks
• References

Were that all, the volume would be a must-read for key actors in society. Yet, at the same time, the authors, with the assistance of Jeffrey Owings and others at NCES, adroitly place this cohort in the context of the changes that have occurred during the cohort's transition from childhood to adulthood. It is easy to forget just how much has changed in the last quarter century. Throughout the volume, the authors remind the reader of the changing context the cohort encountered as it reached key points of transition. Indeed, figure 1 from the report concisely and effectively conveys the diverse and dramatic changes that have occurred since cohort members were born.

Figure 1.—Timeline of milestones in NELS:88 study, average age of cohort members, and selected historical events: 1974–2000

Amidst such change we should not be surprised to find changes in cohort members' experiences as well, but the question, of course, is in what ways did their experiences become affected. For example, one of the most far-ranging transformations of the period was the diffusion of computing technology through virtually all sectors of the economy. This diffusion is reflected by the high proportion of cohort members (over two-thirds) who used a computer at work in 2000. Although use of a computer at work varied by socioeconomic background, nearly half of those from the lowest socioeconomic status (SES) quartile, two-thirds of those from the middle-SES quartiles, and nearly four-fifths of those from the highest-SES quartile used a computer at work. Although this socioeconomic gradient is noteworthy, interestingly, some high-level uses (e.g., writing software) showed no differences by socioeconomic background. Hence, by any measure, computers have transformed the workplaces this cohort occupies. The spread of computer technology is just one example of the dramatic changes that have rolled through societysince the mid-1970s. Coming of Age in the 1990s does an excellent job of keeping the reader aware of how such changes may make a difference.

Amidst such change, however, there are some notable stabilities. Socioeconomic gaps remain substantial in many respects. For example, we learn that nearly 60 percent of students from the highest-SES quartile obtained a bachelor's or higher degree in 2000. In contrast, only 24 percent of students in the middle quartiles obtained a bachelor's or higher degree. Note that these students in the middle quartiles arguably come from middle-class backgrounds. What they are not is the upper middle class, a colloquial term used often in the United States, perhaps to avoid acknowledging that if some are in the middle, then some must be on top. But it is clear that the highest-SES students have markedly better degree attainment prospects than do the middle-SES students. In the same way, the students in the middle-SES quartiles fare far better than those in the lowest-SES quartile; only 7 percent of students from the lowest-SES quartile obtained bachelor's or higher degrees in 2000. In other words, and put crudely, the middle-SES students did about three times better than the lowest-SES students in attaining degrees, and the highest-SES students did about 2.5 times better than the middle-SES students in attaining degrees. In sum, socioeconomic differentials are extremely large.

This is an important set of findings to put before the public, and Coming of Age in the 1990s accomplishes that important task. Further research will be needed to ascertain what these socioeconomic-linked differences mean. If these and other socioeconomic differences are large enough, consistent enough, and robust enough, they may support the theory of effectively maintained inequality (EMI) (Lucas 2001). EMI contends that when there are quantitative differences in a good, the socioeconomically advantaged will use their resources to obtain more of the good. An example of a quantitative difference might be years of schooling; under EMI, we would expect those of advantaged socioeconomic backgrounds to obtain more schooling. And EMI contends that when there are qualitative differences in the good, the socioeconomically advantaged will use their resources to obtain what is qualitatively better. An example of a qualitative difference is provided by tracking: empirical research indicates that students in different educational locations will be exposed to qualitatively different instruction (e.g., Gamoran 1993). We would expect the socioeconomically advantaged to use their resources to obtain the qualitatively better locations, in this case, college preparatory placements in secondary school. Ethnographic evidence is consistent with this expectation (e.g., Useem 1992).

Owing to the ability to trade qualitatively better goods for quantitatively more goods down the road, these processes serve to effectively maintain inequality, perhaps even when access to a good is equalized. Once access to a good is equalized (e.g., access to high school), the socioeconomically advantaged seek out and obtain qualitatively better goods at that level. And those qualitative advantages can be "cashed in" for more of other goods later. In studying the role of socioeconomic background on educational attainment, EMI helped to interpret changes in the effect of socioeconomic background in making several transitions through secondary school on through college entry. This work was done using the High School and Beyond cohort (Lucas 2001). The same kind of process may also be operating in the more recent cohort of 1988 eighth-graders. Whether EMI or some other explanation makes sense of the astounding gross socioeconomic differentials remains for future analysis to discover, but it is clear that the socioeconomic differentials shown in Coming of Age in the 1990s cry out for both continuing analysis and, more important, a policy response.

The foregoing should make clear that Coming of Age in the 1990s is an indispensable volume, and nothing below alters that assessment. Yet it is not all that it could be, nor all that the wider public might need it to be, because of several factors. The analysis is very effective at showing how the changing context must be considered in interpreting the attainments of the cohort. In the same way, the youth-to-adult transition has several key moments, which are, in principle, available, due to the longitudinal design of NELS:88. It would have been helpful to see whether a key interim moment, picked up in earlier waves of data collection, played a big role in the attainments observed in 2000. But this might have been very difficult to do, in part because of the very complexity of the NELS:88 data set. Not only were there several waves of data, with some unavoidable attrition owing to the longitudinal design, but earlier waves were subsampled (to contain cost) and freshened, to make the sample nationally representative at different grades. This subsampling and freshening greatly complicates efforts to use more than two waves in any analysis. The complications arise because the set of cases common across more than two waves may not generalize to any easily identifiable population. Perhaps one reason the snapshot contains only two waves of data is that analyzing more than two waves is just too difficult and too complicated, even for those most closely connected to the data collection. If so, it will be important for future data collection to be designed so that future cohorts to be studied will allow analysts to combine the waves so as to sketch the unfolding experiences of the cohorts in a straightforward way.

A second reason Coming of Age in the 1990s cannot be all the wider public needs it to be is that a wave of data collection would need to be conducted in 2004 or 2005 to answer many questions posed in the base year. In earlier waves, students and young adults often were asked about their plans and expectations. Students were asked to think about a time far into the future—age 30. For 14-year-olds, such a far-off time—a time further away from them than their own infancy—may be difficult to concretely assess. Yet it was the time frame for which expectations and aspirations were ascertained, and it is a reasonable age to select. The 2000 wave, however, assessed the accomplishments and life-course transitions of young adults at a modal age of 26. If we learn anything from Coming of Age in the 1990s, it is that the transitions between ages 14 and 26 were a complex affair for many members of the cohort. The fact that we see that complexity with only two waves of data collection is testimony to the strength of the report. Had the authors analyzed additional intermediate waves, their analysis would only reveal even more complexity.

However, the analytic implication of the complexity is that it is unlikely that one can ascertain whether members of the cohort met the aspirations they set for themselves when in middle school by considering their achievements 4 years before the "deadline" for the realization of those aspirations. Instead, another wave of data collection, measuring accomplishments and attitudes at a modal age of 30, is essential. Absent such a data collection effort, Coming of Age in the 1990s will be our last look at this pivotal cohort. Unfortunately, with that last look, many of the questions that motivated the original investment will remain unanswered. That is not the fault of the authors of the report, but it remains a regrettable circumstance nevertheless.

Coming of Age in the 1990s is an essential volume. It draws on data akin to the Census Bureau's decadal effort to provide a snapshot of the geographic, familial, and socioeconomic location of the nation's inhabitants, with the added complication of connecting observed respondents to "prior" locations. The report is on a par with State of the Union: America in the 1990s—a two-volume analysis of 1990 Census data, a work prepared by more than a dozen analysts across the country—in its scope and depth (Farley 1995). Coming of Age in the 1990s is an illuminating effort—a success. Clearly, before proposing any policy or engaging in any analysis, it will be necessary to check this report to determine what the general tendency has been; whether that tendency variesby important factors such as prior achievement, geographic location, or socioeconomic status; and to locate the experience of youth in the dramatically changing national and international context.

Farley, R (Ed.). (1995). State of the Union: America in the 1990s: Vol. 1: Economic Trends. New York: Russell Sage Foundation.

Farley, R. (Ed.). (1995). State of the Union: America in the 1990s: Vol. 2: Social Trends. New York: Russell Sage Foundation.

Gamoran, A. (1993). Alternative Uses of Ability-Grouping in Secondary Schools: Can We Bring High-Quality Instruction to Low-Ability Classes? American Journal of Education, 102: 1-22.

Lucas, S.R. (2001). Effectively Maintained Inequality: Education Transitions, Track Mobility, and Social Background Effects. American Journal of Sociology, 106: 1642-1690.

Useem, E. (1992). Middle Schools and Math Groups: Parents' Involvement in Children's Placement. Sociology of Education, 65: 263-279.