Your current web browser is outdated. For best viewing experience, please consider upgrading to the latest version.


Send a question or comment using the form below. This message may be routed through support staff.

Email Article

Main Error Mesage Here
More detailed message would go here to provide context for the user and how to proceed
Main Error Mesage Here
More detailed message would go here to provide context for the user and how to proceed

Manhattan Institute

Close Nav
Share this report on Close

New York City's Specialized High Schools: Not the Only Game in Town


New York City's Specialized High Schools: Not the Only Game in Town

April 30, 2019
Urban PolicyNYCEducation
EducationPre K-12

Executive Summary

By law, a student’s admission to one of New York City’s eight elite high schools is determined by his or her score on the Specialized High School Admissions Test (SHSAT). Only a very small number of black and Hispanic students make the cut, a fact that has led Mayor Bill de Blasio to engineer what he regards as a more equitable racial distribution of the widely sought seats in these schools. This will mean that some students, mostly Asian, will lose the seats they would otherwise have earned. Opponents of the mayor’s plan argue that the status quo is color-blind and merit-based and should largely be left alone.

However, the actual effect of these changes for the students who will gain entry and those who lose out can’t be considered in isolation. This report considers the mayor’s plan for the specialized high schools in a citywide context.

Among the findings:

  • There is a clear hierarchy of high schools operated by the New York City Department of Education. Those schools that admit students with the highest average eighth-grade composite state test scores include the city’s specialized high schools as well as other high-performing schools, many with a more racially diverse student body.
  • The eight high schools that use the SHSAT have the highest student attendance (96%), SAT scores (1429 combined verbal and math), and college-readiness rate (99%). However, the next 50 to 75 high schools (in terms of the achievement level of their entering classes) also post high outcomes: attendance rates of 90%–98%; combined average SAT scores of 1000–1400; and college-readiness rates of 80%–95%.
  • The high schools whose students’ eighth-grade composite state test scores were in the two lowest quintiles do not show good outcomes. Attendance data indicate that their students are disengaging from their education in larger numbers; more than 40% are deemed chronically absent. While their graduation rates are reasonable, only a small percentage enroll in college.

A key question about the specialized high schools is whether their students’ impressive outcomes are a function of the schools’ programs and practices or are due to the characteristics of the students they serve. The research on this question is not definitive. Nevertheless, a rigorous study of selective schools in Boston and New York came down on the side of the latter, finding that “while the exam school students in our samples clearly have relatively good outcomes, most of these students would likely have done well without the benefit of an exam school education.”[1]

If more and better educational opportunities for black and Hispanic students are the ultimate goal of the mayor’s plan, his focus on the top eight schools may be nearsighted. They are not the only game in town. Program and curriculum upgrades in the city’s other high-performing high schools might be more effective.



Entry to New York City’s eight specialized high schools is based on the Specialized High School Admissions Test (SHSAT). Very few black and Hispanic middle school students (relative to the share of those groups within the overall school system) score high enough to be offered a seat. Mayor Bill de Blasio regards this as inequitable and wants to change the system.[2]

The mayor’s proposal has two parts. The first part is in place, though it is currently the subject of a legal challenge. Beginning with students entering high school in September 2019 (applications are already in), 20% of the seats will be set aside for students who come through the city’s long-standing Discovery program (see sidebar, The Discovery Program and Racial Diversity) but who just missed the SHSAT cutoff, are classified as disadvantaged, and attend a high-poverty middle school.

The Discovery Program and Racial Diversity

A 1971 state law stipulated that admission to New York City’s specialized high schools would be “solely and exclusively by taking a competitive, objective and scholastic achievement examination.” The statute also established the Discovery Program “to give disadvantaged students of demonstrated high potential an opportunity to try the special high school program without in any manner interfering with the academic level of those schools. A student may be considered for the Discovery Program provided the student: (1) be one of those who takes the regular entrance examination but scores below the cut-off score (2) is certified by his local school as disadvantaged (3) is recommended by his local school as having high potential for the special high school program and (4) attends and then passes a summer preparatory program administered by the special high school. All students recommended for such a Discovery Program are to be arranged on a list according to their entrance examination scores, in descending order, from the highest to the lowest.


Each special high school will then consider candidates in turn, starting at the top of the list for that school.”


Mayor de Blasio increased the number of students admitted to the specialized high schools through Discovery from 58 in 2014 to 250 in 2018.* Yet there was no change in the schools’ racial makeup. More Asian students benefited from Discovery than did black or Hispanic students.


For the coming school year, 2019–20, the de Blasio administration again increased the number of seats set aside for Discovery students but changed the rules. Formerly, students were eligible for Discovery if they had high enough achievement levels and were considered high-needs (low-income). Now, their individual economic status doesn’t matter. Instead, they will be eligible only if they attend a high-poverty middle school. Thus, a low-income student from a middle school that is not predominantly low-income will not be eligible.


*Alex Zimmerman, “New Data Show How Few Black and Hispanic Students Benefit from New York City’s Specialized High School Diversity Program,” Chalkbeat, Aug. 14, 2018

The second part of de Blasio’s plan is to phase out the SHSAT in favor of admissions based on class ranking in seventh grade, which is, in turn, a function of a student’s composite score on the New York State Assessments in English language arts (ELA) and math and seventh-grade course grades. At the end of three years, the SHSAT would be eliminated and 90% of the slots in the selective schools would be filled by students in the top 7% of each public and charter school in the city. The remaining 10% of seats would be reserved for applicants from nonpublic schools and students who are newly entering New York City.

The mayor’s plan has provoked controversy. Yet this controversy has largely ignored a fundamental feature of the city’s public school system: the eight specialized high schools are a small part of a large, complex ecosystem of high schools, many of which successfully prepare their racially diverse students for college. To date, de Blasio has stubbornly demanded that the focus be on the eight specialized high schools alone. This narrow focus on the symbolically rich schools such as Stuyvesant, Bronx Science, and Brooklyn Tech tends to create the impression that these are the only decent high schools in town. That is hardly the case.

This report presents up-to-date information about the other successful high schools, the city’s entire public high school sector, and the test scores of the students who attend them. Here, the admissions criteria to the city’s high schools are important to understand—it has long been known that student outcomes in high school can be accurately predicted by the previous academic achievements of their students.[3] This is critical to thinking about whether students will benefit under the mayor’s proposed changes.

Choice and Screening in New York City’s Public High Schools

New York City’s Department of Education (DOE), America’s largest public school system, operates 419 high schools that enroll more than 276,000 students.[4] Unlike most public high schools in the U.S., few city high schools are geographically zoned or designed to serve all high school–age students in a particular area.

Geographical zoning used to be the rule for New York City, although there were always some high schools that served students from across the city. The specialized high schools were one such group; vocational high schools (which serve students from across the city or borough in which they were located) are another. But the number of geographically zoned high schools declined in the 2000s, as Mayor Michael Bloomberg’s administration aggressively closed low-performing, large neighborhood high schools and replaced them, often with smaller, theme-based schools.[5]

Today, students wishing to attend a public high school must apply through a citywide application system that attempts to place them in their most desired school, consistent with that school’s admissions policy. Late each fall, every eighth-grader from a district, charter, or private school submits a form listing up to 12 programs—such as “science and math,” “humanities and interdisciplinary,” or “health professions” within schools that they wish to apply to—in rank order of their preference. There are 700-plus programs offered across the city’s 400-plus high schools.[6] DOE’s admissions algorithm compares students’ priority rankings with the parameters of programs’ admissions criteria and seeks to offer students entry to the program that they most desire.

Nine high schools do not follow this system. Three of them—Stuyvesant, Bronx High School of Science, and Brooklyn Technical High School—are required by state law to admit students based on the results of the SHSAT. Five others—High School of American Studies at Lehman College; Brooklyn Latin; High School for Math, Science and Engineering at City College of New York; Queens High School for Science at York College; and Staten Island Technical High School—use the SHSAT as a matter of local policy. In addition to these eight specialized high schools, the Fiorello H. LaGuardia High School of Music & Art and Performing Arts is required by state law to admit students based on submission of a portfolio or an audition, depending on the talent area.

In recent years, much of the debate around New York City high schools has focused on observable racial differences across schools—in terms of program availability,[7] access to high-performing schools,[8] and student outcomes.[9] The first step in evaluating these differences is to consider basic data on student test scores.

For decades, scores on the ELA and math tests administered to students in grades 3–8 in New York City’s district schools lagged the rest of the state. Though changes in the state testing program make long-term comparisons difficult, district public schools have shown steady improvement in the years spanning the Bloomberg and de Blasio administrations. In 2006, the city’s performance was 11 percentage points below the state average in reading and 9 percentage points lower in math.[10] In 2018, the city’s traditional public schools scored 1.5 percentage points higher than the rest of the state in ELA and 1.8 points lower in math.[11]

The performance of New York City’s black and Hispanic students on the statewide tests has also been steadily improving.[12] The city’s Hispanic students do better than Hispanic students in the rest of the state in both ELA and math, while black students do better than black students in the rest of the state in ELA and equivalent to black students in the rest of the state in math.[13] Nevertheless, the citywide performance of black and Hispanic students lags 30–40 percentage points behind the city’s Asian and white students.[14]

Given the attention being paid to racial differences in high school admissions, it is worth considering the performance of the city’s students at the conclusion of seventh grade; state tests administered at that point are the last before students must submit their high school applications in the fall of eighth grade (Figures 1 and 2).

New York State ELA and math tests categorize results by proficiency levels. Students at levels 1 and 2 are not considered proficient, students at level 3 are proficient, and students at level 4 excel. Overall, 71% of Asian seventh-graders were considered proficient or above (levels 3 and 4) in math, as were 63% of white students, 26% of Hispanic students, and 21% of black students. For ELA, 65% of Asians were proficient or above, as were 64% of white students, 31% of Hispanic students, and 29% of black students.

Level 4 results may be more relevant to the discussion of admission to the city’s most selective high schools. In math, 47% of Asian students attained that level, as did 37% of white students, 9% of Hispanic students, and 6% of black students. In ELA, 31% of Asians, 29% of whites, 7% of Hispanics, and 6% of blacks attained level 4.

Expressed another way, of all the students attaining level 4 in math in grade 7 in New York City’s district (DOE-run) schools, 44% are Asian, 30% are white, 19% are Hispanic, and 7% are black (Figure 3). The results for ELA are similar. While black and Hispanic students greatly outnumber Asian and white students in district schools, the opposite is true when considering students attaining level 4 on state exams in grade 7. In math, 5,596 Asian students scored at level 4, as did 3,862 white students, 2,372 Hispanic students, and 904 black students.

The number of seventh-grade students in New York City’s charter schools is much smaller, but the test scores of their black and Hispanic students are much higher than in district schools. In math, black seventh-graders in charter schools were more than three times as likely (22.6%) as black seventh-graders in DOE schools (6.6%) to attain level 4. In ELA, black students in charters outperformed their counterparts in DOE schools by a smaller margin—8.5% at level 4, compared with 6.6% in DOE schools.

The characteristics of students upon entry to high school are also important to note. Using 2017–18 DOE data, I analyzed all city high schools based on the test scores of their incoming ninth-graders (composite ELA + math scores from grade 8). I then split them into five groups (or quintiles) by population (each group to consist of approximately 55,000 total students): group 1 had the highest average grade 8 scores, and group 5 had the lowest. The eight specialized high schools were then broken out from group 1 (which now includes 40,000 students) to present them as a separate group, 1A (15,540 students). The groups show clear differences in the composition of the student bodies in terms of socioeconomic and racial characteristics (Figure 4).

The eight specialized high schools serve a student population that differs markedly from other district high schools. The racial differences have been widely discussed—10.3% of the students in the specialized high schools are black or Hispanic and 62.1% are Asian. The much higher academic achievement of the students in the specialized high schools is evident in their average eighth-grade combined state test scores: a full point above that of the next group of schools and 2.5 points above the 5.9 overall district school average on a seven-point scale (minimum = 2; maximum = 9).

The students in the specialized high schools have other different characteristics. Very few (1.3%) are identified as having special needs, compared with the city average of 16.7%. There are also relatively fewer students in temporary housing (homeless, living in a shelter, or “doubled up”) than in the other groups of schools. Nevertheless, 45.6% of specialized high school students are classified as having economic needs, based on an index that considers whether a student’s family has been on public assistance in the last four years and whether they live in a census tract with median income below the poverty level.[15]

The 44 schools in group 1 also serve students with eighth-grade achievement well above the citywide average (1.4 points on the seven-point scale). These vary more from the specialized high schools than they do from the citywide averages on socioeconomic and racial characteristics: 11.8% are deemed to have special needs, 17% are black, 25.1% are Hispanic, 29.2% are Asian, and 24.6% are white.

Schools in the middle groupings (2 and 3) still have significant representation of Asian and white students; they also have a share of special-needs students slightly above the citywide average.

The 252 schools in groups 4 and 5 serve 110,000 students with average eighth-grade achievement below or well below the citywide average. The average student enters high school with test scores below proficiency level; 80%–90% are black or Hispanic; similar portions are from low-income families; and about 20% have special needs. Students residing in temporary housing are concentrated in group 5.

Other measures show the differences in student populations across the city’s high schools (Figure 5). Students in the specialized high schools show very high attendance (96.1% in high school), and relatively few (8.9%) are deemed chronically absent (attendance <90%). Their average scores on Regents Exams in English and algebra are 94.1 and 88.6, respectively.[16]

On average, students in group 1 also show high student attendance (94.3%), though 14.3% are deemed chronically absent. Average Regents Exam test scores are 10 to 13 points lower than those in the specialized high schools but are still well above the city average.

In groups 4 and 5, 42.9 % and 47.2% of the students are chronically absent. The average exam score in algebra is below passing. In English, the average exam score is just above passing in group 4, but below passing in group 5.

Graduation rates (Figure 6) are relatively high in New York City’s district high schools, compared with historical trends—82.8% today across all schools, up from about 50% prior to 2002. Students at the specialized high schools graduate at a 98.7% rate. The graduation rate for students in groups 2 and 3 is also high, 95.7% and 89.3%, respectively. In the lower three tiers, graduation rates range from 81.4% to 71.1 %, though some students will earn diplomas in their fifth or sixth year after entering high school.

The value of a high school diploma has come under scrutiny in recent decades, and another measure has emerged, “college readiness”—the percentage of all students who entered ninth grade in a particular year who meet established thresholds on the Regents Exams, SAT, ACT, or CUNY (City University of New York) Assessment four years later. This indicator matches the graduation rate of students in the specialized high schools, as it should; the schools in group 1 also show a high college-readiness score of 89.1%. Below group 1, college readiness begins to drop off. In group 5, the schools serving those least prepared for high school, only a third of the students attain college readiness within four years. SAT scores of students at the specialized high schools averaged a combined score of 1429, which is 250 points higher than the average for students in group 1 schools. Students in group 3 high schools averaged a combined SAT score of 1053; the average declined to 861 in group 5.

Graduates of New York City’s specialized high schools are much more likely than other graduates to enroll in out-of-state colleges, private colleges in New York State, or public colleges (other than in the CUNY system) in New York State. Overall, 90.9% of specialized high school graduates were enrolled in some type of college six months after graduation, and 9.1% were enrolled in the military, vocational training, AmeriCorps, or City Year Volunteer Corps.

Graduates from high schools in group 1 were enrolled in college at a rate of 82.6%, and they were more likely to be enrolled in CUNY two-year (9.8%) or four-year (30.5%) colleges than were graduates of the specialized high schools. Some 14% of group 1 graduates went on to private colleges in New York State, and 12.6% went on to out-of-state colleges; 15.7% went to public colleges elsewhere in New York State.

Fewer graduates (72.4%) from high schools in group 2 enrolled in college, and they were most likely to enroll in CUNY; 16% enrolled in its two-year colleges, and 29.7% enrolled in its four-year schools. Only 5.6% of graduates from group 2 schools went to out-of-state colleges, while 9.5% went to private colleges in New York State and 11.6% went to New York State public colleges other than CUNY.

In the lowest two tiers of high schools, 48.6% of graduates were enrolled in any type of college six months after graduation. Of these graduates, two-thirds enrolled in CUNY colleges; 35% enrolled in two-year community colleges; and 31% in four-year bachelor’s-degree-granting colleges. Only 5% of groups 1 and 2 high school graduates were enrolled in colleges outside New York State.

We do not have access to data allowing us to link individual students from city high schools into CUNY, but the university system’s overall retention and completion rates can give us a sense of what their college outcomes of groups 1 and 2 high school graduates might be (Figure 7). As a group, students who enter CUNY with lower than average academic performance are certain to have lower success than the systemwide rates presented below.

While 59% of the students who enter CUNY’s bachelor’s programs earn a degree within six years, fewer than 18% of the students who enter its two-year programs earn a degree within three years, and fewer than 26% do so within four years. Yet the Chronicle of Higher Education, using a study by noted economist Raj Chetty, has placed nine of CUNY’s four-year colleges among the top 15 schools in the U.S., in terms of moving low-income students into the middle class.[17]

Do Academically Selective High Schools Make a Difference?

In a 2012 study, Chester E. Finn, Jr. and Jessica Hockett identified 165 high schools across America that they termed “exam schools”—which emphasize academics, prepare students for college, and base admissions “on such academic criteria as grades and test scores.” These schools were educating about 1% of the nation’s high school students.[18] Based on the results of a survey sent to school leaders and site visits to 11 schools, they reported: “The schools we visited were serious, purposeful places: competitive but supportive, energized yet calm. Behavior problems (save for cheating and plagiarism) were minimal and students attended regularly, often even when ill. The kids wanted to be there and were motivated to succeed.”

Nevertheless, the authors noted: “The selection criteria employed by these schools all but guarantee students who are likely to do well academically, which raises the question of whether the schools’ generally impressive outcomes are caused by what happens inside them—their standards, curricula, teachers, homework—or are largely a function of what the kids bring with them.”

The answer to their question about the effectiveness of selective high schools is not simple to obtain, but assumptions about it are at the heart of the controversy surrounding the admissions policies for New York City’s specialized high schools. Will students enrolled under a modified and enlarged Discovery program benefit from attendance at these schools, or will they achieve the same result as they would have in another school? Might they do worse? What about the students who are admitted based on their class ranking in middle schools after the SHSAT is eliminated?

There is some research that bears on these questions. A team of economists analyzed admissions data and student outcomes for exam schools in Boston and New York in a study designed to tease out the school effects and the peer effects associated with these schools.[19] Examining the outcomes of students who are admitted to these schools after meeting the admissions criteria, as well as a separate group of students who fell just short of the admissions criteria, the authors concluded: “Our estimates suggest that the marked changes in peer characteristics at exam school admissions cutoffs have little causal effect on test scores or college quality.” In other words, “while the exam school students in our samples clearly have relatively good outcomes, most of these students would likely have done well without the benefit of an exam school education.”

Similarly, a team of researchers from the Federal Reserve Bank of Chicago and the Chicago Consortium on School Research found no statistically significant impact on student achievement from attending one of Chicago’s Selective Enrollment High Schools (SEHSs).[20] These schools use an affirmative-action admissions policy “reserving seats for students from low-SES [socioeconomic status] neighborhoods, [making them] the most racially diverse public high schools in the city.” After examining outcomes data using sophisticated methodologies, the authors concluded: “We find that when it comes to test scores, attending a SEHS has no statistically significant impact—not even for students from the most disadvantaged neighborhoods or for the highest achieving low-income students who would be admitted even without place-based affirmative action. Given these findings, SEHSs are not helping to close the achievement gap between low- and high-income students.”

The study did, however, find that the schools had a positive impact on the students’ “perceptions of the high school experience.” Students reported feeling safer in these schools and experiencing more respectful interactions with peers, compared with similar students in nonselective schools.

Additional research by the Chicago Consortium on School Research looked at the effect of attending a higher-performing high school on “students’ academic and nonacademic outcomes, relative to the school they would otherwise attend.”[21] Using four-year graduation rates and average ACT composite scores, the researchers divided Chicago’s public high schools into four groups—selective, top tier, middle tier, and bottom tier. Their findings—quite relevant to the current controversy in New York City—found “mixed, but mostly positive, consequences for attending higher achieving schools.” Further, “[t]he effects of attending higher performing schools depends on the outcome being studied and the type of comparison being made. Getting into a selective school is not the same as getting out of a bottom-tier school to go to a mid-tier school, or attending a strong, nonselective school in terms of consequences for students’ experiences in high school, their achievement, or their educational attainment.”

Attendance at “a highly selective school does not give students an academic edge,” according to the study, “at least on academic outcomes: Test scores and coursework are no better, and students’ GPAs are actually somewhat lower if they attend a selective school than a nonselective one.” However, “[t]he most positive effects are observed with strong nonselective schools; students who attend these schools, rather than mid-tier or bottom-tier schools [emphasis added], do see benefits with a number of academic and nonacademic dimensions, including test scores.” They also show “benefits in terms of college enrollment, college selectivity, and college persistence.”

New York City’s Independent Budget Office (IBO) reported the potential impact of de Blasio’s proposal to eliminate the SHSAT and admit students to the eight specialized high schools based upon class rankings in their middle school, which are based upon a combination of course grades, class rank, and state achievement-test results. Using student-level data, the IBO study simulated the differences between the actual student body admitted to the specialized high schools in 2017–18 and the student body that would have been admitted, had the mayor’s proposal been in place.[22]

It found that the mayor’s proposal would have its intended effect—a much greater number of black and Hispanic students would be admitted, while the number of Asian students receiving admission would be cut in half, and there would be little change in the number of white students admitted. More females would be admitted (63% vs. 50%), as would more students in poverty. At the same time, achievement levels of admitted students on the state assessments in ELA and math would decline, though the average seventh-grade course grades would be higher under the proposed changes. Most dramatically, 10% of the students admitted under the mayor’s post-SHSAT plan would have scored below proficiency in math, compared with 0.2% under the current rules.


Academic outcomes in New York City’s district high schools are clearly related to the academic achievement of students before they leave middle school. Much-discussed differences in the racial and socioeconomic characteristics in different high schools derive from the academic sorting that goes on in citywide high school admissions.

Mayor de Blasio’s priority is to change the racial composition of the student body in the city’s specialized high schools. Under the current rules, their students are unique in terms of their academic achievements prior to high school, their SHSAT scores, their high attendance rates while in high school, and their postgraduate college enrollment. Will a change in the admissions policies of the specialized high schools lead to higher achievements of the students who would gain admission under any new scheme? Research from other cities, though not definitive, suggests that it won’t. The reputation of, and high esteem accorded to, New York City’s most selective schools may be largely a function of the uniquely qualified students who attend them. Other students may not necessarily benefit from admission to them. New York’s specialized high schools, moreover, are not the only game in town for those seeking a quality education. There are, for example, a group of 45 high schools just below the specialized high schools in terms of selectivity. While they serve Asian and white students in higher proportions than the city averages, they do serve a greater percentage of black and Hispanic students than do the specialized high schools.

Of these 45 high schools from this report’s group 1 and 2, the 23 in Figure 8 meet three criteria: at least 40% of their students are black or Hispanic; their students attain a college-readiness index of 85%; and students have very high graduation rates, averaging almost 98%. These schools are not “top level”—but they are fertile ground for curriculum and program upgrades to raise their students to the highest levels of achievement.

High schools in the lowest two quintiles (in terms of their students’ entering test scores) serve a population that is almost exclusively black and Hispanic. Some graduate with academic achievements that might prepare them for college; most do not. It is not unreasonable to ask, in light of the mayor’s desire to increase the educational opportunities for the city’s black and Hispanic students, whether more attention should be given to raising the standards of these schools, and perhaps closing more of them. It is also not unreasonable to ask if the school system, as well as the state, might offer more programs aimed at career and workforce preparation for students whose middle school records indicate that they face very long odds in attaining college preparedness in high school.

It is certainly within the mayor’s authority and responsibility to raise questions about fairness within the school system. But his very narrow focus on the city’s eight specialized high schools ignores the opportunities for minority students by expanding and improving those high schools whose racially diverse students are already performing at high levels.


See endnotes in PDF