People are up in arms about Boston’s exam schools. Traditionalists argue the three—Boston Latin School, Boston Latin Academy, and the John D. O’Bryant School of Mathematics and Science—should keep the same rigid admissions requirements they’ve had for decades: a mix of a test scores and students’ GPAs. Reformers want new measures for admittance, measures they hope will diversify the schools’ student populations.
What we should be asking, however, is should we have exam schools at all?
Selective public high schools are a rare phenomenon. There are about 165 nationwide—out of more than 24,000—and they are concentrated mostly in Eastern states. Some are nationally known, including Boston Latin School (founded in 1635, it claims to be America’s oldest school), New York City’s Bronx High School of Science, and Fairfax County, Virginia’s Thomas Jefferson High School for Science and Technology. Schools like these are controversial, with battles raging over the perception that they are “elitist,” creating a two-tier society of the well-educated and the poorly-educated—a divide that often falls along racial and economic lines.
The real problem with selective schools is this: They don’t do much good for the kids who attend them. But they deeply harm the children left behind.
To the first point: Selective schools have little real value. Our belief that they do is attributed to what researchers at MIT called the “elite illusion.” They compared two groups of students in Boston and New York: One group had scores just above the cutoff to get into an exam school; the other was just slightly below. If the schools actually educated kids better then there should have been differences in how the two groups performed on things like standardized tests and college admissions. The researchers found none.
Looking at this study as well as others from schools around the country, Susan M. Dynarski, professor of public policy, education, and economics at the University of Michigan, concluded there was “precisely zero effect of the exam schools on college attendance, college selectivity, and college graduation.” And in particular, she called out Boston, saying its exam schools had “zero effect on test scores, including the SAT and PSAT.”
This may seem counterintuitive, especially to parents like me—both of my daughters went to exam schools. These schools typically do better in the aggregate than other city schools on a host of measures, including overall SAT scores and graduation rates. So they have to be better, right?
Not at all. The exam schools appear better because of what academics call selection bias. It’s not the schools that are so great—it’s their students who are great. “Schools can look like they have a large effect on student outcomes,” observed researchers who studied Chicago’s selective schools, “while these apparent successes should actually be attributed to the students themselves.” That’s not to say schools never matter. But students who are already high academic achievers tend to remain that way no matter where they are educated.
It’s not only that exam schools don’t help. By putting all of the most academically advanced students in just a handful of schools, the non-exam schools are far worse off.
This happens in a number of ways. One is simply that children tend to respond to peer pressure. This year, there are 21,095 children in grades 7 through 12 of the Boston public schools. Fully 28 percent of them—5,894—are in the three exam schools (which begin in Grade 7). And what happens to everyone else? “Students are influenced by their peers’ attitudes, achievements, and choices, and if low-achievement students get concentrated in certain schools, a culture of low achievement/low ambitions develops there,” argue researchers Magnus Bygren and Erik Rosenqvist.
Furthermore, there’s the phenomenon of “expectation bias.” By depriving schools of their highest achievers, teachers at the non-exam schools can draw the false conclusion that those who remain are not as smart and not as capable. And a large body of research demonstrates that these kinds of assumptions can, in effect, be self-fulfilling—they “have a causal impact on students’ educational attainment,” say Nicholas Papageorge and Seth Gershenson in a Brookings Institute paper. Not surprisingly, such expectation bias also cuts across racial lines, with data showing that teachers in general tend to expect poorer performance from their Black students than their white students.
Aside from these concerns, admissions to exam schools are deeply unfair. Children from poorer circumstances with parents less able to afford private tutors and exam prep courses tend not to get in—an issue that also cuts across racial lines. Moreover, using achievement when children are only in Grade 6 as a marker of whether they are worthy of an exam school ignores the wildly different ways kids develop intellectually—some more quickly than others—as well as the variety of different learning styles each of us has. Exams and GPAs focus on only one kind of intelligence—the “book smart” kind—to the detriment of the host of others (in 1983, psychologist Howard Gardner proposed eight: logical-mathematical, linguistic, spatial, musical, bodily-kinesthetic, intrapersonal, interpersonal, and naturalistic).
Granted, the exam schools are the Boston public school system’s shining lights. They’re exceptionally well-resourced (check out Boston Latin School’s Harry V. Keefe Library, for instance, or its roughly 30 Advanced Placement offerings, far more than the typical non-exam school), with multiple clubs and activities that offer their lucky students experiences they can’t find in regular schools. But the costs of that approach are high and the solution seems clear: Instead of just three shining lights, we should illuminate them all.
Tom Keane is a frequent contributor to the Globe Magazine. Send comments to firstname.lastname@example.org.