全球智库信息集成服务系统

By some measures, educational attainment is the highest it has ever been. National high school graduation rates have risen every year since they were first collected in the 2010-11 school year by the U.S. Department of Education.

Yet, ensuring that all students complete high school remains an elusive goal. Approximately 1 million high school students each year fail to earn a diploma or its equivalent.¹ This number translates to a national school dropout rate of 5.9 percent, but for certain subgroups of students, specifically black and Hispanic students, it is much higher—6.5 and 9.2 percent, respectively.

Under federal law, high schools with graduation rates that are less than 67 percent or meet other criteria for low performance are subject to intensive improvement strategies. This requirement also applies to what the law defines as “alternative education campuses” (AECs), schools that states have established to serve the unique needs of students who are at risk of dropping out or who have re-engaged in school.

Federal policy allows states to use the same, or different, measures to hold these schools accountable for their performance as other public schools. However, there is a knowledge gap when it comes to understanding how students navigate the alternative school experience and how effectively the federally required school performance measures assess these schools.

Recent analyses show that without meaningful accountability, traditional school districts may push struggling students into low-quality alternative schools.² The analysis shared in this report suggests that measures used to hold these schools accountable may over-identify failure and under-identify success. School accountability systems better designed to measure the nuances of student experience in these schools would provide critically needed insights.

This paper provides the groundwork to design such measures. The recommendations aim to improve researchers’, practitioners’, and policymakers’ ability to conduct much-needed investigation into the experience of students in alternative schools, while striking a balance between accurate measurement and rigorous expectations. It presents two options for states to consider in developing federal accountability systems for alternative schools.

First, states could forego using metrics in federal law, as these calculations inadequately determine performance for students who are poorly served by traditional schools. This report will demonstrate this inadequacy by looking at the experience of the New York City public schools and proposing metrics that better capture this experience. States interested in developing customized metrics should conduct similar analyses of their student outcome data for alternative schools.

Second, states could use the 67 percent graduation rate metric to identify schools as low-performing, while using recommendations presented in this report—including a graduation rate index, credit accumulation, and attendance—to gauge progress to exit low-performing status.

In this report, the authors review the characteristics of alternative schools on a national level, profile the students who attend them, and outline the legal history of these schools. The report also explores a series of options to more effectively measure the performance of alternative high schools. While meaningful school accountability systems include the examination of a wide array of data and the implementation of supports to ensure continuous improvement in all schools,³ this report focuses exclusively on measuring school progress. Data on school progress is a necessary first step toward designing broader systems of support, improvement, and resource allocation. The proposed school progress metrics fall into three areas: graduation rate; academic proficiency; and school quality and student success. These proposed metrics draw upon efforts in alternative high schools in the New York City Department of Education

For more than 20 years, standards-based accountability has been a cornerstone of education policy in the United States. This effort has included setting the same, rigorous academic standards for all students, and holding schools, districts, and states accountable for students meeting those standards. Yet, this has not always had the intended consequences. Data show that some states lowered their proficiency standards in reading and math so that more students could count as proficient under the No Child Left Behind Act.⁴ The updated law, the Every Student Succeeds Act (ESSA) of 2015, requires that proficiency standards mean that students meeting them do not need remedial education when in college. ESSA’s requirement intends to set the expectation that all students should be prepared for the next step when they receive a high school diploma.

Similarly, for more than 10 years the nation has used a four-year cohort graduation rate for federal accountability purposes, setting the expectation that students should graduate from high school within four years. These consistent expectations for high schools are intended to address inequities in access to a high-quality education and inequities in outcomes for different groups of students. Based on how students progress through school, however, the traditional measures of educational progress do not adequately measure how well alternative education campuses, specifically alternative high schools, are serving their unique population of students.

Traditional student performance measures assume a cohort of students making regular progress—that is to say, students enter high school and four years later they graduate. In between, teachers and tests judge their progress. Most students follow this linear path meeting traditional performance measures along the way and graduating at the end of four years. However, some students fall behind for myriad reasons.

To understand this dynamic more deeply, Eskolta School Research and Design studied seven years of data for nearly 70,000 New York City public school students whose first year of high school was 2008, tracking whether they had graduated by 2014, six years later. In the equivalent of ninth, 10th, and 11th grades, 7 out of 8 of these students made regular progress. However, one-eighth did not, falling two or more years behind within their first three years of high school.⁵ Of the students who had not fallen behind in their first three years, 88.9 percent graduated by 2014. Of those who did fall behind, however, the chances that they graduated by 2014 were far worse. Most remained in traditional high school settings; of these, just 13.2 percent graduated by 2014, a rate 75.7 percentage points lower than that of students who had not fallen behind. But nearly 2,000 of the students who had fallen behind instead transferred out of traditional high schools and into alternative high schools; of these, 29.9 percent graduated by 2014, more than double the rate of their counterparts in other high schools. Measured against a single statewide standard, this figure appears dismal, but measured against the track record of the students who are entering, it appears outstanding. The reality, however, is that measures like the four-year or six-year graduation rate simply do not do justice to the lived experience of these students.

The vast majority of students who transferred from traditional high schools to alternative high schools did so following experiences that disrupted the linear progress assumed by traditional measures of performance—many of which were traumatic and stigmatizing. In the cohort studied in New York City, for example, this disruption in education is reflected in the fact that more than three-quarters (77.7 percent) of the students who transferred to alternative high schools had been chronically absent the year that they transferred.⁶

For these students, progress may be more erratic. A child experiences trauma—a violent episode, problems with substance abuse, financial hardship, a family illness, depression—after which school performance suddenly and dramatically falters. Dropping out of school or being held back multiple times can itself be experienced as a traumatic event. Students who fall behind may feel that they no longer belong to an academic community or may doubt their ability to succeed, mindsets that foreshadow poor academic performance.⁸

Graduation measures that treat these students as if they are part of a cohort that is proceeding without interruption contradicts their experiences. From a purely statistical standpoint, there is little to no chance that they will graduate in four years. In the cohort studied in New York City, for example, nearly 1 in 5 students (18.3 percent) arrived at an alternative high school after their fourth year of high school. An additional 37.1 percent arrived at an alternative school during their fourth year.⁹ Including these students in a four-year cohort relegates the schools into which they transfer to a dataset that virtually guarantees a “negative” graduation rate, since calculation of graduation rates counts every student enrolled, even if that student arrived just weeks or months before their graduation date. As this report highlights below, students in alternative schools can achieve positive school completion outcomes that are different than those typically counted by accountability systems.

Beginning in 2001 with the No Child Left Behind Act and continuing under ESSA, states could design accountability systems that better measure the unique characteristics of alternative schools. To do so, states must better understand the characteristics of alternative schools and their students.

There is no consensus definition of alternative schools; each state determines its own grade configurations and eligibility criteria.¹⁰ This lack of consensus makes it difficult to ascertain how well these schools serve their students.

For example, exactly what counts as an alternative school varies by the data set. The National Center for Education Statistics (NCES) reports demographic and outcome data for alternative schools, defined as, “A public elementary/secondary school that addresses the needs of students that typically cannot be met in a regular school program. The school provides nontraditional education; serves as an adjunct to a regular school; and falls outside the categories of regular, special education, or vocational education.”¹¹ However, since states decide what counts as an alternative school for federal school accountability purposes, there are discrepancies in calculations of how many of these schools exist, how many students they serve, and what outcomes they generate when comparing state and federal data on these schools.

Evidence of the likely discrepancy between definitions used at the district, state, and federal level can be found in the estimates from NCES placing the number of alternative schools at 6,448 in 2005-06, 10,300 in 2007-08, and 6,293 in 2009-10. Here, we will use the 2014-15 school year data from NCES, which identified about 4,548 alternative high schools and 435,088 students enrolled in these schools nationwide, representing almost 17 percent of all public high schools and almost 3 percent of public high school students.¹²

These national data also indicate that alternative schools are much more likely to experience closure than any other school type. For example, even though alternative schools make up only 6 percent of all schools they make up 23 percent of all school closures.¹³ Closure may be due to low graduation rate or other factors. In fact, some of the same factors that make aligning alternative high school definitions difficult—their small size and mixed grade level offerings, for example—also make understanding their average graduation rates difficult.

While these data do shed light on the characteristics of alternative high schools and students, it is critical that stakeholders reach a more aligned definition. Without such alignment, there will never be an accurate national understanding of how well these schools are serving their students.

The Every Student Succeeds Act requires that all public schools be held accountable for their performance using five types of annual indicators that states must customize. These indicators are to be used for annual school performance identifications and for periodic identification of low-performing schools. ESSA-required indicators for all schools include: academic proficiency; English language proficiency for English learners; and at least one measure of school quality and student success. High schools also use graduation rate as an indicator; elementary and middle schools, meanwhile, use another academic indicator, such as student growth.

In March 2017, the U.S. Department of Education provided a revised template for states to submit their consolidated ESSA state plans. This template gives states the opportunity to describe a distinct methodology for annually identifying school performance for schools for which an accountability determination cannot be made.¹⁴ While not explicitly noted in the state plan template, since states must use their annual performance system to also periodically identify low-performing schools, it logically follows that this alternative methodology could be used for both purposes: annual performance determinations and periodic identification of low-performing schools.

Given federal accountability requirements, the limitations in their usefulness to alternative high schools, and the significant flexibility available in holding them accountable, this report proposes a series of indicators that better measure academic attainment, high school completion, and other measures of what federal law calls “school quality and student success.”¹⁵ These indicators are based on the experience of alternative high schools in New York City and can be useful for states considering developing a unique system of accountability for alternative high schools.

In New York City public schools in recent years, a series of experiments have been underway. In 2006, the school district created a new school progress report card specifically for alternative educational campuses—which New York calls “transfer schools”—and supported a wave of openings of these schools.

The district’s transfer schools were originally defined as those serving high school students who enroll after having previously been enrolled in another high school. The definition has since expanded to include some schools that enroll ninth graders who enter high school already two or more years behind. While students enroll in these alternative high schools for myriad reasons, and such schools have been options in New York for decades, recent attention has been on the students in the district’s transfer schools who are over-age and under-credited, meaning the number of high school credits they have earned given their age is two or more years behind what it should be.

In the past decade, the district has gradually revised and refined accountability metrics that treat high school graduation and academic growth differently for its transfer schools than for its other high schools. Over a similar time, the nonprofit New York Performance Standards Consortium, a coalition of schools across the state, brought together 36 New York City schools, eight of which are transfer schools, to pioneer project-based assessment tasks (PBATs) to measure student academic proficiency.

These efforts are occurring alongside real and meaningful change for students and schools. Currently, the district has more than 50 transfer schools enrolling nearly 15,000 students, a dramatic increase in both number and enrollment over the last 15 years. Alongside the proliferation of transfer schools, the city has seen improvements in its state-reported graduation rates. From 2011 to 2016, when national graduation rates rose a healthy 4 percentage points, in New York City graduation rates outpaced these significantly, with four-year graduation rates rising 7.5 percentage points (from 65.5 percent to 73.0 percent) and six-year rates 5.7 percentage points (from 70.9 percent to 76.6 percent).¹⁷

While graduation rates among transfer schools are lower than that of traditional high schools, data suggest that students attending transfer schools would, in the absence of these schools, have graduated at rates at least 50 percent lower than they did.¹⁸ As referenced earlier, reviewing data on nearly 9,000 students who fell two or more years behind in their first three years in one cohort of New York City public high school students, Eskolta found a graduation rate of 13.2 percent for those who remained on register at a traditional high school, compared to a graduation rate of 29.9 percent for those who instead transferred to an alternative high school.¹⁹ A study by Metis Associates for Good Shepherd Services showed that students who attend transfer schools using a model that intentionally incorporates youth development and counseling practices to reintegrate students into the high school setting are 12.6 percent more likely to graduate than students with similar demographic and academic characteristics.²⁰

Do states have performance indicators that are appropriate to alternative high schools? Effective indicators should relate to key student outcomes, meaningfully differentiate among elements that lead to those outcomes, and incentivize positive action at the school and district level.²¹ The indicators recommended in this report draw upon the pioneering work of New York City to do that. They address four aspects of measures to inform policymakers’ understanding of the alternative high school population:

• How will measures be calculated such that schools and districts are focused on the right outcomes? For students whose success in conventional high schools was hard to discern, cultivating success in new environments requires care and attention from the adults who make these calculations.
• Which students’ data will be attributed to which schools? When students have struggled at one school, then transfer to an alternative education campus, decisions of data attribution become complicated. At what point are students attributed to a school?
• Which students are considered part of the same cohort? This decision enables schools to differentiate among groups of students and, when organized meaningfully, is a better incentive to action. In conventional high schools, this is simply defined as the students who entered ninth grade at the same time. In schools that accept the clear majority of their students after ninth grade, defining a cohort becomes more complicated.
• Against what benchmark will schools’ results be compared? When a high school mostly enrolls students with predicted outcomes that are far weaker than average, that school is what statisticians call a “biased” sample. How can these outcomes be differentiated to consider the characteristics of the sample of students that are in an alternative high school?

This report focuses on presenting new measures for graduation rate, academic proficiency, and what ESSA calls “school quality and student success” generally addressing the four aspects detailed above.

Graduation index

Because of the students that they enroll, alternative high schools are chronically identified as underperforming due to low four-year graduation rates. In 2017, for example, of the 51 New York City transfer schools with sufficient data to report four-year graduation rates, only two schools would have met the 67 percent graduation rate needed to avoid the underperforming designation under ESSA.²² Put simply, a four-year or even extended-year cohort rate would over-identify these schools as low graduation rate schools. To better identify alternative high schools’ rates of school completion and to group students in a way that brings meaningful differentiation and incentivizes action, states should create an alternative high school graduation index that contains the following elements:

• Calculation. Include all regular high school diploma graduates, not only those who graduate within four, five, or six years. Recognize other positive school completion outcomes, verified by documentation from the school leader, on a case-by-case basis.
• Attribution. Attribute student data to schools after an agreed period of continuous enrollment determined by the state.
• Cohort. Use a single-year graduation with a regular high school diploma cohort based on exit or targeted exit dates.
• Comparison. Consider establishing a peer comparison benchmark or subgroups against which graduation outcomes above are compared to measure outcomes. 

Calculation: Inclusion of all graduations and identified school completion outcomes 

Federal guidelines generally emphasize four-year graduation rates. New York’s City’s “transfer graduation rate,” created specifically for its accountability reports for transfer schools, instead includes all students who earn high school diplomas in a given year, regardless of how many years they have attended high school. This approach has the advantage of counting diploma-earning graduates who would not be counted in a traditional four-year graduation cohort. In New York City in 2017, for example, 1,852 students who earned their diplomas after more than four years were counted in the transfer school graduation rate, making the transfer school graduation rate of 51 percent slightly more than double the 24 percent four-year graduation rate for transfer schools.²³

In 2012, New York City added to its progress reports for all high schools a postsecondary success metric that credits schools not only for graduates, but also for students’ enrollment in two- or four-year colleges, vocational programs, or public service programs. Nearly all of these data come from the National Student Clearinghouse or from the City University of New York. To supplement these data for students not included in these sources, principals can attest to other postsecondary outcomes, provided that they retain documentation on file that is periodically audited by the district. The supplementary information is a step toward generating goodwill and investment from principals of the schools being held accountable, in that it takes a broad view of success, an important element of the success of an accountability system.

Attribution: Assigning students to schools

Which school is held accountable for a student’s data once he or she exits school? For example, if a student drops out of one high school after three years, then enrolls in an alternative high school and drops out after three days, which school is held accountable for that subsequent dropout? If accountability immediately transfers with the student, the alternative high school in this example would be held accountable for that student’s second dropout—even though that school had minimal opportunity to influence the student’s growth. On the other hand, if accountability takes too long to transfer, then the student’s first school remains accountable for a student’s subsequent dropout even after it no longer wields influence on that student’s outcomes. States should identify an agreed period of continuous enrollment after which students are attributed to the school.

Cohort: Inclusion of all graduates 

Nationally, the graduation cohort is calculated as the percentage of all students who entered ninth grade at the same time and graduated four years later. But conventional cohorts can be confusing for schools such as alternative high schools, where virtually no students enrolled when they were in ninth grade. Indeed, most transfer schools in New York City simply do not use the notion of traditional ninth, 10th, 11th, or 12th grade because these do not differentiate among transfer high school students in a way that is meaningful. Furthermore, conventional graduation rates are labeled “four-year” and “six-year” when alternative high schools in fact have far less time than this. In the cohort studied in New York City, as noted earlier, fully 55 percent of students entering their first year at transfer high schools were already in their third year or later of high school.²⁴

With a waiver under No Child Left Behind in 2008, New York State adopted a single-year entry cohort for testing measurements in its transfer schools. Instead of focusing on the year that students entered ninth grade, every student who entered the transfer school in a given school year was included in accountability measures three years later. Further analysis by New York City has since highlighted problems with this approach, particularly in that it overlooked the progress of schools that worked with students who were significantly far behind and therefore needed more than three years to recover learning.²⁵ One way of addressing this issue would be to place students in cohorts based on their credit level at the time of enrollment into the school, such that entering students with the same general number of credits needed to reach graduation are placed in the same cohort. For example, all students who enroll with less than one-quarter of the credits needed to graduate high school would be placed in one cohort while those who enroll with one-quarter to one-half the credits needed to graduate would be placed in another cohort with a different targeted graduation year.

A variation on the single-year entry cohort is a single-year exit cohort. All students who exit the school in a given year are counted in this graduation rate. For instance, if 100 students exit an alternative high school in 2017-18, of whom 60 do so with high school diplomas and 40 do not, then the single-year exit graduation rate for the school’s cohort is 60 percent. While this approach has not been utilized in New York City, it has the advantage of connecting to a set of students who were all enrolled in the school in the year for which the school is held accountable and of counting every student, thereby better meeting the test of incentivizing school-level action. A drawback, however, is that data may be skewed by the length of time students are enrolled. For example, consider a school that largely enrolls older students who are very close to graduation, but also serves some who are not as old and remain enrolled for significant time without graduating. This school would experience a relatively high graduation rate using a single-year exit cohort m