Exam High Schools: Not As Great As We Thought

(Digital Vision)

Exam high schools are generally regarded as a cut above, turning out congressmen, scholars, and all-around high achievers. They account for over half of the top 109 American schools in the U.S. News and World Report best high schools list, and an incredible 20 out of 21 from Newsweek’s list of “public elite.”

But a new study from Will Dobbie and Roland Fryer of Harvard throws cold water on this notion, and calls into question whether the exam schools typically cited for excellence are, well, really all that excellent.

Writing for the National Bureau of Economic Research, Dobbie and Fryer take a fresh look into the measurable achievements of exam school students, specifically focusing on three well-known schools in New York City: Brooklyn Tech, Bronx Science, and Stuyvesant. While attending an exam school might be great for your overall education, and resume, this doesn’t come through in terms of increased test scores or college achievement. Here’s the abstract:

Publicly funded exam schools educate many of the world’s most talented students. These schools typically contain higher achieving peers, more rigorous instruction, and additional resources compared to regular public schools. This paper uses a sharp discontinuity in the admissions process at three prominent exam schools in New York City to provide the first causal estimate of the impact of attending an exam school in the United States on longer term academic outcomes. Attending an exam school increases the rigor of high school courses taken and the probability that a student graduates with an advanced high school degree. Surprisingly, however, attending an exam school has little impact on Scholastic Aptitude Test scores, college enrollment, or college graduation — casting doubt on their ultimate long term impact.

The numbers on college attendance and graduation are by far the most surprising – exam high schools have a lower college attendance and graduation rate compared to other high schools. The study breaks down the numbers according to individual high schools.

Students just eligible for Brooklyn Tech are 2.3 percentage points less likely to graduate from a four year college. Students just eligible for Bronx Science are 0.7 percentage points less likely to graduate, and students just eligible for Stuyvesant are 1.6 percentage points less likely to graduate, though neither estimate is statistically significant.

Though college graduation and the SATs might not show a winning formula, Fryer and Dobbie are careful to point out that the criteria they’re examining don’t paint the whole story.

… without longer-term measures such as income, health, or life satisfaction, it is difficult to fully interpret our results. To the extent that attending an exam school increases social capital in ways that are important for later outcomes that are independent of college enrollment, graduation, or human capital, then there is reason to believe that our conclusions are premature and the true impact of an elite exam school will only be understood with the passage of time.


Chess Piece Face

What the heck is an "exam school"??

Joe Dokes

It is a public school that uses an exam for entrance. Thus, only the "best and brightest" get in. Once in they grind the living s%$t out of the students.

I am not surprised at all of the results of this study. We don't have exam schools per se in Long Beach Ca. but most local high schools have various college prep magnet programs. In 1986 the Long Beach Polytechnic High Schools PACE program required two AP classes for graduating seniors, today it is five. Thus, a high school senior from the PACE program enters college with 15 or more units of college level course work.

On first glance this would be great, but the anecdotal evidence is not encouraging. I've seen numerous students who enter college pre-burnt out. (Colleges even have a term for them, they are called crispies.) Many fail to finish college, and many of those that do, end up majoring in something that has little interest for them, and have not thought through careers they might wish to pursue, so they languish.

A rigorous high school is necessary for true learning. Many "exam school" programs are more interested in f-ing the US News metric than in actually turning out students who are both prepared for college level work and the desire to actually go to college and learn.


Joe Dokes



"Many “exam school” programs are more interested in f-ing the US News metric than in actually turning out students who are both prepared for college level work and the desire to actually go to college and learn."

Sounds like Campbell's Law has struck again.


Hm. Counter-intuitive.

But I like the authors’ caveats: Is college graduation really an appropriate proxy for “good outcomes”? Perhaps people with rigorous high school prep don’t require as many years of college before they leap into the job market. But that hypothesis simply prompts the question, is labor force participation an appropriate proxy for “good outcomes”? Thus the authors suggest studying income, health, and life satisfaction as relevant long-term outcomes. In short, while there may be no agreement about what is the “good outcome,” and there may be no substitute for picking some specific good outcome when attempting to measure the efficacy of education.

As an aside: My college boasted of the high percentage of its graduates that attend graduate school. Does that mean that I attended a good college (because grad schools recognize the quality of its graduates) or a bad one (because none of the grads could find jobs, and therefore had to pursue a real education elsewhere)?



I went to a DC area magnet school, and out of a graduating class of 400, we had one student not going on to a 4-year college. It seems to me that the school didn't affect the odds of any given student going to college, since the students would have been top-tier students at the public high-school they attended (also, a sizable minority would have gone to private high-schools had they not been admitted to the magnet school).

The main benefit for me was twofold: no longer being the big fish in the small pond academically, and the ability to form social groups. The latter one was particularly impactful, as "smart kids" was not really a specific enough label, so the cliques formed along different lines there.

Enter your name

The caveat is important. Bill Gates, Larry Ellison, Mark Zuckerberg, and Steve Jobs are all "college dropout", but nobody would call them failures.

The social side of these schools is important for bright kids. It helps them redefine themselves in ways other than the stereotypical genius-nerd. Also, it helps them discover that they really *aren't* better and smarter than everyone else, which is an important life lesson that they will not learn in a mixed-ability classroom. Consequently, I'd like to see a study that measures life satisfaction or general relationship success (not just romantic success).


This is not surprising at all. I would expect a similar outcome for a study of those who "just barely" got accepted to MIT, Caltech, etc.

These borderline students might well be better off being at or near the top of a decent school, rather than struggling at a more demanding and more competitive school.


Interesting point.

The post talks about the overall rate at which students attend and graduate from college: “The numbers on college attendance and graduation are by far the most surprising – exam high schools have a lower college attendance and graduation rate compared to other high schools.”

But in support of this statement, the post then cites language that apparently does not discuss the overall attributes of students from these schools, but rather the attributes of the marginal student at these schools: “Students just eligible for Brooklyn Tech are 2.3 percentage points less likely to graduate from a four year college. Students just eligible for Bronx Science are 0.7 percentage points less likely to graduate, and students just eligible for Stuyvesant are 1.6 percentage points less likely to graduate….”

Maybe someone could clear this up?



I wasn't actually clear on that either - I thought it was probably referring to the marginal students, but the language is ambiguous.

Regardless, I think those marginal students are going to be the bulk of the underachievers in college. It may also be that those marginal students are marginally more likely to go to tougher college than if they hadn't gone to an exam school - and graduate slightly less often because they aimed too high.


This seems silly, the real questions are do the kids who go to exam schools get into more elite colleges then a similar cohort of peers? Does that translate into higher lifetime earnings? Obviously, those two factors are why parents want their kids to go to exam schools.


Also keep in mind that the kids who get into "exam" schools (or magnet schools are they are more commonly know) are by definition very smart and/or movitated and therefore likely to be successful irrespective of which high school they end up attending.


Congressmen and scholars are considered high achievers?

Maybe conspicuous achievers, but not necessarily high achievers. It would make sense to me that students who are drawn to exam schools would also be drawn to other forms of conspicuous achievements.


It's a self-fulfilling prophecy. If I am creating a football team, but only allow the fastest, biggest, most skilled football players on my team, yeah, I'm going to have better results than the school that gets the leftovers. Same with academics. When you filter it such that only the best and brightest get into your school, then, yes, you're going to have better results.

Of course, the same thing goes with Harvard and Yale, and the such. When you are a marquee name, you can be highly selective. The best-performing students in the world want in. After all, a degree from Harvard means big bucks often.

But that doesn't mean that these kids are getting a better education. In fact, they may be getting a quite mediocre one, yet because they are good students, they do well. Put those same students in Podunk Junction State College...same thing. They are good students and will excel wherever they go. But it makes much more sense to excel at Harvard than at PJ State.

At the same time, if you put DECENT students in a great college, I believe they can achieve more than they would have otherwise. But if you put poor students just about anywhere, you'll still get poor results.

Going to an exam school is no guarantee that the student will be better educated (though that might be the case), but it likely a guarantee that your child's chances for successfully reaching the college of his/her choice is upped.


Glen Craig

I graduated from one of these schools, Brooklyn Tech, and I have to say using college graduation and attendance doesn't seem to prove anything about the success of the school.

After high school I attended CUNY's Baruch college. Many of the courses I took were not challenging. Why? I had already done work like that in high school. After four years of high school it's not very exciting taking the same courses again in college. I'm one of those that hasn't finished college

You need to look long-term at what the students have done after high school rather than their college attendance and graduation rates. This is the problem with education in general - we only want to see numbers that seem to indicate success rather than see if we are actually educating people. Instead of aiming for well-rounded people we look for well-scored people. Sigh.


Considering that standardized tests are created to test what students are supposed to be taught in a regular high school, and that colleges base their curriculum on continuing off of the public high school curriculum, both of which are sensible practices, it makes sense that students in an "advanced" program would not test better- their advanced learning is likely not being tested.

I went to a small, rural high school in Tennessee, and I am certain my scores were above the average of any school- I know my ACT was above the average for students at Vanderbilt University by two points. Does that mean Exam Schools are a waste? I don't know- would I have scored that one more point possible on the Math section and gotten that one more overall point on the test if I had been at a "better" school? Or was I, as a student in a school that was teaching what it was supposed to teach, fully equipped for the test because nothing "advanced" was tested?



With SAT scores only predicting 15-20% of first year college gpa, I wouldn't be surprised that high school entrance exams are substantially equivalent. If so, then eventual college graduation by exam school graduates should not be much different than the general population. Graduation from college depends on adjustments to academic and social conditions very different from high school. One of the things that educational researchers are discovering is that transitions from elementary to secondary and secondary to post-secondary can be difficult for students. Just when they become accustomed to a style of pedagogy and a social environment, it changes and knocks some of them back. A system based on chronological age does not allow for differing rates of maturity that enable students to make transitions between levels easily. Only recently with the investigation of the K-through-16 year experience has this problem started to be addressed. Up to now college faculty usually just have complained that new freshmen weren't prepared for college work.



I think the Freakonomics authors need to find more solid studies before they give them this kind of airing. Studies "just eligible" for this column should be rejected.

Suzanne Brown

My late husband was educated in and taught for 42 years in independent schools. The last 10 years of his teaching career was in a very unusual school - a totally independent school, governed by a Board of Trustees, which also served the community as the public high school - the state paid for these students to attend as there was no public high school in this community. The test scores for this school in a poor area of a rural state was 3rd in the state, only behind the two school districts in the wealthiest suburbs of the state, many filled with the children of doctors and lawyers. How did they do this? Small class size and high expectations for all. Did all of these students go on to college? No, but most did, even if the colleges were not necessarily the most competitive. The parents loved that their tax dollars were getting an independent school education for their children.
My husband always said that the super-bright children are what he called "teacher-proof" children - they will do well in the lousiest school on earth. They are genetically very, very bright, and they are self-driven and highly motivated. He believed that independent schools (and would have said the same about exam magnet schools) best served the bright but not brilliant student. That student can easily fall through the cracks in a large public high school. They're not brilliant enough to get the teacher's attention, and are well-behaved enough to not get that attention! But being in a school with excellent teachers who truly love what they do, and being with a peer group with high expectations will help those students achieve more than they would have in a typical large public high school. I know that was true for my own daughter.
My eldest grandson is presently attending a math & science exam/magnet school and it has been an eye-opener for him. At his elementary school, he was considered one of the two or three smartest students in the school; in him magnet middle school, he's probably in the middle of the pack. He has come up against the work ethic of the children of asian doctors and while he still gets A's in all of his classes, he does not get 100% on all tests as these other children do. The top 30 in the school were put in an even more advanced program - another exam program. He did not make the cut. He's learned that just getting an A will not make one the smartest child in the class - he needs to strive for greater perfection. We will see if he has the work ethic to do so - his parents expect him to get A's but not 100%.
The one thing that is ALWAYS left out of any conversation about student achievement is the role of the parents. Some parents barely know where their children go to school; others go over every single piece of homework nightly and insist on re-doing anything that is not perfect. Most fall somewhere in between. One can poor all the money on earth into special educational programs to try to advance achievement, but if the genes for brightness are not there and there is not considerable parental evolvement and support, the finest teachers on earth will not succeed. Bright children with bright parents who value education and are supportive will succeed; others will not.