Event, status, and cohort dropout rates each provide a different perspective on the student dropout population. The includes definitions and data for each type of dropout rate in order to provide a detailed profile of dropouts in the United States. High school graduation and completion rates complete the profile of high school outcomes for young adults in the United States.
Event rates calculated using the October 1996 CPS data measure the proportion of students who dropped out between October 1995 and October 1996.10 These dropouts are 15- through 24-year-olds who were enrolled in high school in October of 1995, but had not completed high school and were not enrolled in grades 10 through 12 a year later.11 By October 1996, 5 out of every 100 young adults (5 percent) enrolled in high school in October 1995 left high school without successfully completing a high school program (table 1 and C1).12
Over the past quarter century, annual estimates of the event dropout rate have fluctuated between 4.0 and 6.7 percent (figure 1 and table C3).13 There was a downward trend in the estimated percentage of current year high school dropouts in the 1980s and early 1990s.14 However, changes in survey methodology complicate the analysis of recent estimates. The annual event dropout rates declined between 1978 and 1986, when a change in the editing procedures for the school enrollment items may have affected reporting.15
More recently, additional changes in data collection and estimation procedures coincided with an apparent increase in the rates from 1991 through 1995.16 In fact, the 1995 event dropout rate of 5.7 percent was significantly higher than the rates registered for all but one year between 1986 and 1993.17 By comparison, although the 1996 rate of 5.0 percent is not significantly lower than the 1995 rate, it is also not significantly higher than the rates registered between 1986 and 1993. (table C3)
The percentage of young adults who left school each year without successfully completing a high school program decreased from 1972 through 1986, but the rates over the subsequent years have not shown a consistent pattern. Although there have been year-to-year fluctuations, the net effect leads to the conclusion that the percentage of students dropping out each year has neither increased nor decreased since the late 1980s.
The 1996 CPS data are consistent with earlier reports of a strong association between race-ethnicity and dropping out of school. In particular, cohort studies of national longitudinal data for American high school students, such as the High School and Beyond survey sponsored by NCES, show that Hispanics and blacks are at greater risk of dropping out than whites, with Hispanics at a greater risk of dropping out than either white or black students.18 More recently, analyses of data from the NCES National Education Longitudinal Study, and analyses reported by the White House Panel on Hispanic Dropouts also confirm these patterns.19
Data from the October 1996 CPS repeat this pattern, showing an event dropout rate of 9.0 percent for Hispanic students, higher than the rate of 4.1 percent for white students.20 The estimated rate for black students (6.7 percent) falls between the rates for Hispanics and whites, but the differences are not significant (table 1).21
Race-ethnicity is only one of a number of closely linked factors that mediate a student's decision to drop out of school. A number of other factors that more than likely play an important role in this decision are highly correlated with an individual's racial and ethnic background. Just to name a few, examples include socioeconomic background, the ability to communicate in English, and geographic region of residence. Analyses of all of the specific interactions among intervening variables that mediate the dropout decision are beyond the scope of this report. Instead, this report reviews some of the primary factors that are associated with higher event dropout rates.22
The Current Population Survey includes family income data that can be used to provide information about the impact of socioeconomic background and home environment on the decisions of these young adults to drop out. Of course the range of factors that affect young adults' life decisions extend beyond the economic conditions associated with family income; however, in the absence of additional measures, family income serves as a good indicator for the other social and economic factors that are likely to have an impact on a young adult's decision to stay in school. In 1996, 11.1 percent of students from families in the lowest 20 percent of the income distribution dropped out of high school; by way of comparison, 5.1 percent in the middle 60 percent of the income distribution dropped out, and 2.1 percent of students from families with incomes in the top 20 percent dropped out (table 1).
Comparable data spanning the last quarter century provide strong evidence that income, and more than likely the complex of social factors affected by income, makes a difference in the dropout decision process. The annual event dropout rates for students with family incomes in the lowest 20 percent of the family income distribution range from 4.5 to 11 times the dropout rates recorded for students with family incomes in the top 20 percent of the family income distribution (figure 2). A comparison of the annual dropout rates at each income level suggests that there is a larger gap between the dropout rates for students in the lowest versus the middle income group than there is between the rates for students in the middle and highest income groups. A more detailed examination of these data shows that, on average, the gap between the dropout rates in the highest and lowest income groups is 11.5 percentage points. Two-thirds of this gap is accounted for by differences between students in the lowest compared to the middle income group (a difference, on average, of 7.8 percentage points); while the remaining one-third (on average, 3.7 percentage points) is due to the gap in dropout rates between students in the middle and highest income groups.
In October of 1996, only 1 out of every 10 youths ages 15 through 24 enrolled in school was over age 18, but dropouts from this older group of students accounted for 1 out of every 4 high school dropouts in 1996. Thus, students who pursue a high school program beyond the traditional ages are at an increased risk of dropping out of school (table 2).
While the event dropout rates for younger enrollees are substantially lower (for example, only 3.5 percent for 15- and 16-year-olds and 3.4 percent for 17-year-olds), it is important to understand that 43 percent of all young adults who left school between October of 1995 and October of 1996 were ages 15, 16, and 17 in October of 1996. These youths left school short of a projected normal school completion. Understanding why these younger students, with presumably fewer career options, choose to leave school early is an important issue to resolve.
For the past four years, the Common Core of Data (CCD) universe collection at NCES has included a dropout component in the agency level nonfiscal data collection. Currently NCES, through the National Cooperative for Elementary and Secondary Statistics and the CCD collection, is working with states and school districts to develop this national database of public school dropout rates. The number of states participating with consistent data definitions and collection procedures has increased from 14 states plus the District of Columbia for school year 1991-92 to 17 states for school year 1992-93 and most recently to 29 states for school year 1994-95. Once all states are participating fully in this data collection, event data for sex, race-ethnicity, and for grades 7 through 12 will be aggregated at the state and national levels. In the 1995-96 school year collection, 44 states plus the District of Columbia submitted data to CCD for dropouts from the 1994-95 school year. Data from 29 states meet the quality and comparability levels necessary for publishing state level estimates that support valid cross-state comparisons. The middle case, or median, of the dropout rates for these states is 4.6 percent registered by Delaware, Indiana, and Rhode Island, with rates ranging from 2.5 percent in North Dakota to 10.3 percent in Nevada (table 3).
The cumulative effect of hundreds of thousands of young adults leaving school each year without successfully completing a high school program translates into several million young adults who are out of school but lacking a high school credential. Each year over the last decade this number has exceeded 3 million.23 In October of 1996 there were 3.6 million 16- through 24-year-olds who were not enrolled in a high school program and had not completed high school (table 4). Overall, 11.1 percent of the 32.5 million 16- through 24-year-olds in the U.S. in 1996 were in this group.
Although only 7.3 percent of white young adults ages 16 through 24 were out of school without a high school credential in 1996, they account for 1.6 million of the 3.6 million dropouts (table 5). An additional 1.3 million Hispanic dropouts account for 29.4 percent of the Hispanic young adults in this age group. Moreover, 13 percent of black 16- through 24-year-olds add another 0.6 million young adults to the dropout count.
The 1996 status dropout rate of 11.1 percent is lower than the 1995 rate of 12.0 percent, and continues an overall pattern of decline in the status rate. This 1995 to 1996 decrease is reflected in the status dropout rates for white youths, which decreased from 8.6 percent in 1995 to 7.3 percent in 1996, but is not repeated in the case of blacks or Hispanics, whose dropout rates showed no significant changes between 1995 and 199624 (figure 3). In recent years, annual fluctuations in these data have made it difficult to interpret short-term comparisons. However, a time series analysis based on the annual data provides a framework for describing longer term patterns of change. Over the past 25 years, there has been an overall pattern of decline that, on average, amounts to a change of 0.13 percent per year.
Over the past quarter century, the status dropout rates for white young adults have persisted at levels lower than the rates observed for either black or Hispanic young adults (figure 3). However, over the time period, the percentages of white and black young adults out of school without a high school credential have declined by about 40 percent in each group. Since the dropout rates for black young adults have been higher than those for white young adults, the comparable rates of change have resulted in a narrowing of the gap between the rates for blacks and whites.
Hispanic young adults in the United States have not shared in this improvement. Over the last 25 years, close to one-third of the 16- through 24-year-old Hispanics in the United States were reported as out of school and lacking a high school credential.25 The educational gap between Hispanic young adults compared to their black and white peers is made worse by the fact that taken as a group, Hispanic young adults without a high school credential have completed less schooling than black and white young adults in the same situation. For example, one-third of the Hispanic young adults in this group of dropouts have less than a 9th-grade education and one-half have less than a 10th-grade education (table 6). Comparable estimates for whites and blacks show that about one-tenth of the dropouts in each group have less than a 9th-grade education and about one-quarter have less than a 10th-grade education.
Earlier reports that the higher Hispanic dropout rates are in part attributable to high dropout rates among Hispanic immigrants are substantiated with data from 1996 (table 7).26 In fact, the status dropout rate of 44.1 percent for Hispanic 16- through 24-year-olds born outside the 50 states and the District of Columbia is double the rates of 16.7 percent registered for Hispanic youths with at least one parent born in the United States and 22 percent registered for Hispanic youths with both parents born in the United States. While these dropout rates for Hispanic youths born in the United States are lower than the dropout rate for foreign-born Hispanic youths, they are still higher than the dropout rates registered for black or white young adults, leaving a larger share of the group of Hispanic young adults ill-prepared to compete for skilled or technical jobs in today's economy.
Data from 1995 show that over half of the foreign-born Hispanic youths who were counted as dropouts never enrolled in a U.S. school, and 80 percent of these young adults were reported as either speaking English "not well" or "not at all."27 Some of the young Hispanic immigrants who do not enroll in school in the U.S. may have entered the U.S. beyond what is considered "normal" high school age, and some may have come to the U.S. in search of employment rather than education. But the data also suggest that language may be a barrier to participation in U.S. schools. Regardless of the reasons that resulted in a large proportion of Hispanic young adults not having a high school credential, the impact is the same; whether they were born in the 50 states and the District of Columbia or elsewhere and whether or not they enrolled in U.S. schools, these young adults do not have the basic level of education that is thought to be essential in today's economy.
The patterns described for event dropout rates and income levels are repeated in the status dropout rates. In 1996, there was almost a 20-percentage-point gap between status dropout rates for the lowest and highest income groups (table 8). Youths from families with the lowest incomes were eight times as likely as their peers from families with high incomes to be out of school without a high school credential.
The clear differences in the distribution of dropout rates across racial-ethnic groups and clear differences in the distribution of dropout rates across income levels lead to questions about the relationship between racial-ethnic group membership, income levels, and dropout rates. As reported in past years, within each racial-ethnic group, status dropout rates are lowest at the highest income levels and highest at the lowest income levels.
Despite these patterns, differences between racial-ethnic groups persist within income levels (table 8). Although Hispanic young adults from families with low and middle incomes are more likely to drop out than Hispanic youth from families with high incomes, Hispanic young adults at each income level are more likely to drop out than white and black youths at the same income levels. Comparisons of dropout rates for white and black youths at each income level show that at the middle and high income levels the dropout rates for black and white youths are comparable. However, black young adults at the low income level do not fare as well as their white peers. White and black youths in families with low incomes are more likely to drop out than their peers in families with higher incomes, but the risk is greater for black youths (21.9 percent for blacks versus 13.9 percent for whites).
The 5.7 percentage-point difference between the rates for blacks and whites is, in part, due to the differences evident in the dropout rates for white and black youths at the lowest income levels; but the size of this gap is also driven by differences in the population distribution across the income levels. Relatively more white than black youths live in families in the highest income group (28.6 percent versus 9.7 percent), while a larger share of black than white youths live in families in the lowest income group (35.7 percent versus 13.3 percent) (table 9). As a result, a larger portion of black youths are at the increased risk of dropping out observed in the low income group (21.9 percent for blacks) and a larger portion of white youths experience the decreased risk of dropping out observed in the high income group (2.0 percent for whites).
Historically, geographic regions have been another area of interest in efforts to understand patterns and trends in dropout rates. Similar to findings in 1995, the high status dropout rates of 13.0 percent in the South and 13.9 percent in the West are at least one and one-half times the rates of 8.3 percent in the Northeast and 7.7 percent in the Midwest (table 5). When these dropout rates are reviewed across regions for each racial-ethnic group, the dropout rates for Hispanics exceed the national dropout rates in each region, but there is no clear pattern for Hispanics across regions (table 10). The rates for black youths are on a par with the national average in each region except the West. Black youths in the West have dropout rates lower than the national average and lower than the dropout rates for black youths in each other region of the country. The rates for white youths are lower than the national averages in each region. But the rates for white youths in the South are higher than the rates experienced by white youths in each of the other regions. The South is the only region in which the dropout rate for white youths exceeds the national dropout rate for white youths (10.0 percent versus 7.3 percent).
Taken together, these patterns suggest that the higher rates evident in the West are partially driven by the high dropout rate for Hispanic youths.28 Even though the dropout rate for white youths in the South is lower than the dropout rates for black and Hispanic youths in the South, the fact that it is the highest regional dropout rate for whites suggests that the greatest challenge to further lowering the white dropout rate lies in the South.
Longitudinal studies follow the experiences that a cohort of students share as they progress through school. This type of study provides an opportunity to examine in more detail questions about who drops out, the life circumstances of dropouts, the factors that influence the decisions to drop out of high school, and the experiences young adults encounter after leaving school. The NELS:88 started with the cohort of students who were in the eighth grade in the 1987-88 school year.29 Subsequently, these students were re-interviewed at two-year intervals through 1994.
The cohort dropout rates for the eighth-grade class of 1988 show that by the spring of 1992, 10.8 percent of the 1988 cohort of eighth graders were out of school and had not completed a high school program (table 11). Some of these dropouts completed a high school program over the following summer, so that by August of 1992 the size of this group was 10.1 percent. By August of 1994, only 7.2 percent of the cohort remained as dropouts.
Analysis of the outcomes experienced by these dropouts shows that relative to their peers who completed high school, they were less likely to participate in postsecondary education; on average, they earned lower incomes; and they were also more likely to make early transitions into adult roles-to have children or marry or live in marriage-like arrangements.30
Comparisons can also be drawn across cohorts measured at the same point in their school careers but in different years. The NELS:88 also included a nationally representative sample of sophomores in 1990; these students were re-interviewed in 1992 and 1994. Comparable data were collected for sophomores in 1980 in the HS&B study; these students were re-interviewed in 1982 and 1984.
A comparison of cohort dropout rates from the 1980 and 1990 sophomore classes shows that 9.9 percent of the students who were sophomores in 1980 were high school dropouts by August of the 1981-82 school year (table 12).31 For the sophomore class of 1990, the cohort dropout rate was lower, with 5.6 percent of the students who were counted as sophomores in 1990 counted as dropouts by August of the 1991-92 school year. 32 This amounts to a 43 percent reduction in the sophomore to senior dropout rate over the decade.33
Decreases in dropouts rates were widespread, with a number of different groups of students sharing in the decline. Dropout rates decreased for both male and female students, for white, black, and Hispanic students, for students living in intact families and non-intact families, and for students with children of their own living in their household. At the same time, students in poverty and with relatively poor academic achievement seem to be left untouched by the combination of factors that led to lower dropout rates over the 10-year period. Sophomores with these characteristics dropped out at comparable rates in 1980 and 1990.34
 Specifically, the numerator of the event rate for 1996 was the number of persons 15 through 24 years old surveyed in 1996 who were enrolled in high school in October of 1995, were not enrolled in October of 1996, and also did not complete high school (i.e., had not received a high school diploma or an equivalency certificate) between October 1995 and October 1996. The denominator of the event rate is the sum of the dropouts (i.e., the numerator) and the number of all persons 15 through 24 years old who attended grades 10 through 12 and are still enrolled or graduated or completed high school.
 The statistical significance of these comparisons was assessed with Student's t-test with a Bonferroni correction for multiple comparisons. For a full discussion of the statistical methods used in this report, see appendix B. All comparisons in this report are statistically significant at the a = 0.05 level.
 The wording of the educational attainment of the CPS was changed in 1992. Furthermore, data since 1994 may reflect changes in CPS due to newly instituted computer-assisted interviewing, and also may reflect the change from 1980 census-based estimates to 1990 census-based estimates, with adjustment for undercount.
 See R. Ekstron, M. Goertz, J. Pollack and D. Rock, "Who Drops Out of High School and Why? Findings from a National Study," in School Dropouts: Patterns and Policies, ed. G. Natriello (New York: Teachers College Press, 1987), 52-69. For dropout data using the National Education Longitudinal Study, see the NCES publication, Dropout Rates in the United States: 1994.
 For more in-depth coverage on the interaction of race-ethnicity with other factors, the interested reader is referred to G. Natriello, Ed., School Dropouts: Patterns and Policies (New York: Teachers College Press, 1987). For an excellent ethnographic depiction of these factors at work, see M. Fine, Framing Dropouts (New York: State University of New York Press, 1991).
 For data from 1985-1991, see P. Kaufman and M. Frase, Dropout Rates in the United States: 1989 (Washington D.C.: , U.S. Department of Education, NCES 90-659); P. Kaufman, M. McMillen, and D. Bradby, Dropout Rates in the United States: 1991 (Washington, D.C.: , U.S. Department of Education, NCES 92-129).
 See for example, F . Bennici and W. Strang, An Analysis of Language Minority and Limited English Proficient Students from NELS:88 (Washington, D.C.: U.S. Department of Education, Office of Bilingual Education and Minority Language Affairs, August 1995); W. Strang, M. Winglee, and J. Stunkard, Characteristics of Secondary-School-Age Language Minority and Limited English Proficient Youth (Washington, D.C.: U.S. Department of Education, 1993); and P. Kaufman and M. McMillen, Dropout Rates in the United States: 1990 (Washington, D.C.: U.S. Department of Education, , NCES 91-053).
 See tables 16 and 20, M. McMillen, P. Kaufman, and S. Klein. Dropout Rates in the United States: 1995 (Washington, D.C.: U.S. Department of Education, , NCES 97-473). The English-speaking ability is based on the reports of a household respondent rather than reports from each individual in the household. These data on the ability in speak English are limited to young adults who were reported as speaking Spanish at home.
 For more information, see S. Ingels, S. Abraham, K. Kasinski, R. Karr, B. Spencer, M. Frankel, and J. Owings, NELS:88 Base Year Data File User's Manuals (Washington, D.C.: Student Component: March 1990, NCES 90-494; Parent Component: March 1990, NCES 90-466; School Component: March 1990, NCES 90-482; and Teacher Component, March 1990, NCES 90-484); and B. Spencer, M. Frankel, S. Ingels, K. Rasinski and R. Tourangeau, NELS:88 Base Year Sample Design Report (Washington, D.C.: NCES 90-463, 1990).
 Previous analyses of the HS&B data from the spring 1982 follow-up counted students who had enrolled in alternative programs to prepare for a high school equivalency test or had completed high school by an alternative means as dropouts. See S.M. Barro and A Kolstad, Who Drops Out of High School? Findings from High School and Beyond (1987); and A. Pallas, "School Dropouts in the United States" (issue paper, U.S. Department of Education, , 1987). The analysis presented here treats these youths as students or completers.
 In both HS&B and NELS:88, a subset of students who were not considered capable of completing the questionnaire were deemed ineligible for participation in the study. Inasmuch as no attempt was made to identify and include data from students deemed ineligible in the 1980 HS&B cohort, analyses that compare NELS:88 sophomores with HS&B sophomores do not include data reflecting the experiences of the ineligible students in NELS:88. The option for school coordinators to determine some students ineligible led to the exclusion of an unknown number of language minority and limited English proficient students in HS&B. In NELS:88 however, a Spanish-language questionnaire was administered to those members of the sophomore cohort who preferred to take this version of the questionnaire.