Photo from Bigstock.

As part of her Year of the High School initiative, DC Public Schools Chancellor Kaya Henderson is expanding Advanced Placement offerings at all DCPS high schools. But at most high-poverty DC high schools, few if any students earn passing grades on AP exams.

Starting this year, DCPS is raising the minimum number of AP courses each high school must offer from four to six. Next year, all high schools will be required to offer at least eight AP courses.

The expansion of AP in DC is part of a nationwide trend, fueled by the idea that all students benefit from taking the ostensibly rigorous, college-level classes regardless of how well prepared they are.

The Washington Post’s Jay Mathews, a leading proponent of the AP-for-all theory, publishes an annual ranking of US high schools based largely on how many AP tests they administer per graduating student. Although Mathews’ methodology and assumptions have drawn criticism, his ranking has spurred much of the AP growth.

Nationally, AP participation rates have more than doubled in the past decade, with 2.5 million students taking at least one AP exam in 2015. But as the number of AP test-takers has expanded to include many more low-income and minority students, the failure rate has grown even more rapidly.

AP exams are graded on a scale of 1 to 5, and the College Board, which administers the exam, considers 3 to be a passing score, enabling a student to earn college credit. (Many universities give credit only for scores of 4 or 5.)

Nationwide, about 60% of all test takers scored a 3 or better on at least one exam. But the pass rate for African-American students was just half the rate for white students.

In DCPS overall, the proportion of exams on which a student earned a 3 or above has gone up from 27% in 2010 to 33% in 2015. But those figures, provided by DCPS, don’t reveal much about how students at each school are performing. Students at some schools may be taking several AP exams, doing well on all of them.

A DCPS spokesperson, Michelle Lerner, declined to release school-by-school pass rates, saying that some AP classes are small enough that individual students could be identified. But a retired DCPS teacher, Erich Martel, has calculated school-level scores on the basis of information he received from an internal DCPS source. The data lists scores for all AP tests taken at each DCPS high school in 2012 and 2013.

At some schools, almost all tests get a score of 1

In 2013, according to Martel, the overall pass rate for DCPS was just under 31%. But that rate drops to lower than 10% if you exclude relatively affluent Wilson High School in Ward 3 and three selective schools—Banneker, Duke Ellington School for the Arts, and School Without Walls.

At four high-poverty DCPS schools—Dunbar, Ballou, Cardozo, and the now-closed Spingarn—none of the tests received a passing score. At Coolidge, H.D. Woodson, and Anacostia, the pass rate was less than 4%.

Overall, almost 46% of tests taken by DCPS students got the lowest score possible, a 1. But again, if you exclude Wilson and the three selective schools, almost 70% got that score. At Spingarn, all 24 tests received a 1, and at Dunbar, 49 out of 52 did.

Mathews and other advocates of AP expansion argue that students benefit from the experience of taking AP classes and tests, even if they don’t pass the tests. Some studies have supported that claim, while others have refuted it.

The most recent study concluded that merely taking an AP class, without also taking the test, had no effect on a student’s score on the ACT college entrance exam. Those who took and failed the AP test scored a quarter to half a point higher on the ACT, which is roughly equivalent to the boost a student would get from test prep coaching. (Students who passed the AP test scored from one to four points higher on the ACT, depending on which AP class they took.)

About 95% of DCPS students who take AP classes also take the test, according to DCPS. But the lead author of the recent AP study believes that students benefit not from the three hours spent taking the test but from the studying they put in beforehand. So the real question may be: do the many DCPS students who get 1s on AP tests actually study for them?

No doubt some AP teachers out there could answer that better than I can. But when I volunteered as a tutor for a college-level history class at a high-poverty high school a couple of years ago, I realized that most students in the class lacked the background knowledge and vocabulary to gain even a basic understanding of the texts. And if students can’t understand the material, they can’t study for the test.

An AP score of 1 could mean that a student showed up for the test and just answered questions randomly, or didn’t answer them at all. A study cited by Mathews in support of AP expansion shows benefits for students who get “even a score of 2” on the AP, but says nothing about those who get a 1.

AP-for-all defenders argue that even if students are unprepared for AP classes, they’ll get more out of them than they would out of regular classes where exams are graded, not by an independent entity, but by teachers who may be willing to lower standards. But if the AP material is far above students’ heads, they may not be getting anything out of the classes at all. Perhaps we need a third alternative: classes that are both rigorous and accessible to the students who are taking them.

While building a pipeline, keep AP classes small

“We believe that at every school there are students at AP level,” says DCPS’s Lerner. That may be true, but at high-poverty schools even those students probably need a good deal of support to do well. And unless classes are small, they won’t get it.

Lerner says DCPS schools are required to offer the minimum number of AP classes even if only a few students enroll. But will school administrators resist the temptation to herd large numbers of students into classes they’re not prepared for, as they seem to have done in the past?

Even if administrators keep AP classes small, without additional funds the result may be that other classes get larger. And the non-AP students may get inferior teachers, since schools generally assign their best teachers to AP classes.

Lerner says the district is “setting up a pipeline” for its AP classes and expects enrollment to grow in the future. That makes sense: if you want low-income students to be prepared for AP classes in high school, you need to start laying the foundation in kindergarten, if not before.

But increasing the number of AP classes now at all schools makes sense only if DCPS ensures the classes are limited to students who can actually get something out of them.

Cross-posted at DC Eduphile.

The original version of this post said that the scores compiled by Erich Martel were not broken down by AP subject. In fact, they include both aggregate AP scores for each school and subject scores.

Natalie Wexler is a DC education journalist and blogger. She chairs the board of The Writing Revolution and serves on the Urban Teachers DC Regional Leadership Council, and she has been a volunteer reading and writing tutor in high-poverty DC Public Schools.