Laura Fuchs has been teaching Advanced Placement U.S. government and politics at H.D. Woodson High School in Washington for six of her 10 years there. She is in her early 30s, wears her hair pulled back in a bun and has a no-nonsense way of dealing with her students. But that apparent sternness belies a genuine love of teaching and a deep well of patience, two qualities that have prepared her for teaching a college-level course at a school like Woodson.

Many students in Washington go to charter schools or to one of seven magnet high schools in the district, which admit students based on specific eligibility criteria. Woodson, serving around 600 students in the second-poorest ward in the city, is one of nine high schools in Washington with open enrollment. In 2016, it had a graduation rate of 76 percent — up from 53 percent in 2012. Last year only 1 percent of its students met math standards on national standardized tests, and 4 percent met reading standards. Woodson is among the lowest-performing schools in the city, and as at many of Washington’s public schools, 100 percent of the students receive free or subsidized lunches. “A lot of students are not at the poverty line; they’re significantly below it,” Fuchs says.

On the day I visited last fall, Fuchs’s class was learning about Social Security. Most of Fuchs’s students were already familiar with the program, but only its disability-insurance component. Many were surprised to learn that it’s also a retirement program. Next, Fuchs brought up the 1950s baby boom and discovered that some of her students didn’t realize that the country’s birthrate is lower today than it was in the past. One student chimed in jokingly: “That’s O.K., these high school students can catch them up!” Fuchs steered them back to the matter at hand. “When it comes to the A.P. exam, you need to know what Social Security is — it’s called an entitlement, which means if you qualify according to the law, you receive it,” she explained.

A.P. U.S. government, like the 38 other A.P. courses developed by the College Board, a nonprofit organization, is a difficult class. Students are expected to read college-level textbooks, grasp complicated vocabulary and concepts and spend 30 minutes to an hour each night on homework. At the end of the year is an arduous final exam designed, distributed and graded by the College Board. If students score a 3 or better on a 5-point scale, they typically receive college credit. (Though the College Board does not consider a 1 or a 2 to be a failing grade, they are commonly understood to be — and students receive no credit for them.) As of last year, D.C. Public Schools required that all its high schools offer at least eight Advanced Placement classes.

Until recently, this fact alone would have been considered remarkable. A.P. classes were, for years, primarily taught in wealthier school districts. But over the last decade, the program has grown rapidly. In 2006, 1.3 million students took at least one A.P. exam; by 2016, the number had increased to 2.6 million. The total number of tests taken grew during the same time period to 4.7 million from 2.3 million. Much of this growth is due to increased federal funding for A.P. tests and concerted efforts by the College Board to reach low-income and minority students. The organization has a program called “All In,” which identifies lower-income students who might succeed in an A.P. class based on their PSAT scores — the Preliminary SAT, which the College Board also administers — and then reaches out to those students (and their teachers and advisers) to persuade them to take the courses.

From one viewpoint, the expansion has been successful. In 2005, only 6.4 percent of the nation’s high school seniors who took A.P.s were black; that figure increased to 9.5 percent in 2015. Hispanics’ participation grew to 20 percent from 13.4 percent. For low-income students, that figure doubled, to about 30 percent from about 15 percent.

But taking an A.P. class and succeeding at it are two different things. After class let out, Fuchs told me, with a note of frustration in her voice, how few of her students passed the A.P. exam at the end of the year. “I’ve got five to six kids reading on grade level, and three of those don’t show up,” she said. “The rest are significantly below grade level.” Fuchs, whose class meets for 85 minutes daily, works with her students through lunch periods and whenever she has a free moment. She took some of them canvassing out of state on weekends during the 2016 campaign to teach them about real-world politics. She buys them breakfast the morning of the A.P. exam. Many of her students are engaged and passionate. Still, she said, “for six years, the passage rate has always been completely flat.” Usually, only one or two students in her class score a 3 or higher.

Fuchs’s students are part of a broader trend. Nationally, over 70 percent of African-Americans and 57 percent of Hispanics who took an A.P. test in 2016 did not pass. (Over all, the failure rate was 42 percent.) And over the past two decades, although the percentage of students scoring between 2 and 5 remained fairly stable, the percentage of students scoring 1 has grown to 19 percent from 12 percent.

In 2016, at the nine open-enrollment neighborhood high schools in Washington, the passage rates were fairly dismal; at three schools, only one student score a 3 or above, and one had no students pass at all. This failure rate, which is rarely highlighted by the College Board — or the policy makers and legislators who also drive the A.P. expansion — raises questions that are as tangled as any about race, class and education in this country. Critics of the program see the A.P.’s expansion as a boondoggle, with scarce resources being thrown at a program that simply wasn’t designed to address the systemic problems facing public education — at a real cost to these students.

The Advanced Placement program began in 1955, inspired by a fear that American high school students were falling behind the rest of the world, the Soviet Union in particular. By offering elite high school students an opportunity to take college-level classes, the United States could theoretically regain ground it had lost. The program was initially developed with funding from the Ford Foundation and was eventually taken over by the College Board, which had been administering standardized tests since the beginning of the 20th century. As the courses grew in popularity, they soon became a useful assessment tool in college admissions offices, according to Kristin Klopfenstein, director of the Colorado Evaluation and Action Lab and one of the foremost experts on the A.P. program. American schooling standards differed regionally, but, she says, “A.P. scores mean the same thing whether you’re in New York or Louisiana.”

The A.P. program remained a mainstay of affluent, mostly white schools until the 1990s, when parents in lower-income school districts became increasingly concerned about the disparity between the number of A.P. classes offered at their schools and the number in wealthier districts. Rigorous standardized tests, it was thought at the time, could be a means of bridging the achievement gap between richer and poorer schools. In 1999, the A.C.L.U. sued the state of California on behalf of black and Hispanic high school students in Inglewood, who were denied equal access to A.P. courses, saying the state violated the students’ right to an equal education. Inglewood High School in South Los Angeles offered only three A.P. classes, while Beverly Hills High School offered 45 A.P. classes in 14 subjects. The state settled the lawsuit by agreeing to increase access to A.P. classes. In the following years, the U.S. Department of Education’s office of civil rights found that numerous school districts as far-flung as South Orange and Maplewood, N.J., and Lee County, Ala., were not providing equal A.P. access to African-American students.

With expanded access, the A.P. curriculum’s reason for being grew more complex. “A.P. is now being asked to serve multiple purposes in society,” says Chester E. Finn Jr., a former assistant secretary of education under President Reagan who is now a senior fellow at the Hoover Institution. What started as a program for accelerating the education of gifted students is now being used as a means of broadening access to challenging material, Finn says.

Questions about the A.P. program’s purpose are complicated further by the fact that it provides a not-insignificant amount of revenue for the College Board. Of the College Board’s total $916 million in revenue in 2015, $408 million came from fees for the test and instructional materials. (Next year, the test fee will be $94.) The A.P. could become even more important as an income generator in the face of financial and brand challenges for the organization. The College Board also administers the SAT, which has been losing ground in recent years to the ACT. And over the past decade, some elite private high schools have begun dropping the A.P. curriculum in favor of their own homegrown honors programs.

  • The College Board is guilty of promising too much, offering its rigor as a cure for struggling school districts.

Nevertheless, the College Board has effectively persuaded politicians and policy makers that A.P. classes and tests are one of the best ways to raise standards in poorer schools. Federal and state governments cumulatively have spent hundreds of millions of dollars on A.P. tests and teacher training. As of last year, 29 states subsidized A.P. exams, 20 states offered financial incentives like teacher bonuses for strong A.P. scores and 30 states required that A.P. participation or scores be used in measuring school and district performance. In New York City, Mayor Bill de Blasio’s “A.P. for All” initiative, which started last year, promises that by 2021 all city schools will provide students with access to at least five A.P.s, at a cost of $41.5 million annually for teacher training, student tutoring and materials.

“There has been a tremendous push to market for good and bad reasons,” says Theodore O’Neill, who served for 20 years as the University of Chicago’s dean of admissions. “I think the whole accumulation of power in the hands of the College Board that determines what a curriculum should be is suspect in itself.” Diane Ravitch, a former assistant secretary of education appointed by President George H.W. Bush — and a fervent critic of the growth of standardized testing in American education — puts it succinctly: “The A.P. has become a cash cow for the College Board. If it really wants to promote equity, offer the tests for free.”

Trevor Packer, the senior vice president of Advanced Placement and instruction at the College Board, denies that the organization views the A.P. expansion as a road to growth. “No one has ever said: ‘Trevor, we need you to increase the revenue,’ ” he told me. For him, not expanding access to A.P.s would mean a tremendous amount of lost potential. “If I were to look at kids taking too many A.P.s or not taking enough — I’m concerned about both, but one thousand times more concerned about kids not getting access to any A.P.s,” Packer says. “I would rather have a culture where we take risks on giving opportunities to kids.

Packer pointed to the College Board’s own research, which shows that students aren’t discouraged by failing the exams. In fact, students who received a 1 or a 2 on an A.P. exam in 10th grade were significantly more likely to take an A.P. exam later in high school than those who had not taken an A.P. test in 10th grade. “I don’t see how that’s harming anyone,” he says. Packer believes that the numbers actually signify success. “The overall A.P. score hasn’t changed much,” he told me. In 2008, the mean score was 2.85; in 2016, it was 2.87. “We don’t see much cause for concern.”

  • “A lot of money is being spent for students to take the test, and a lot of students are not being successful.”

But to critics, the College Board is guilty of promising too much, offering its rigor as a cure for struggling school districts — something it was never meant to be. In 2015, The Journal of Negro Education, a Howard University publication, released a study of A.P. exam scores, incentives and costs in Texas, Florida and New York. The authors of the article noted that while “the A.P. program has been popularized as one of the most effective strategies used in closing the achievement gap, preparing students for college and careers and gaining admissions to postsecondary institutions,” those claims “by the College Board do not mirror the lived experiences of black and Hispanic student groups.” Or as George W. Moore, an author of the report and an associate professor at Sam Houston State University in Huntsville, Tex., says, “A lot of money is being spent for students to take the test, and a lot of students are not being successful.”

It’s easy to see why expanded access to A.P. courses is such a tempting idea. Take, for example, Anthony Yom, 37, a math teacher at Abraham Lincoln High School, a low-income public school in Los Angeles. Yom teaches calculus AB (the lower of the two levels of A.P. calculus) and makes himself available to his students virtually around the clock — before and after school, at lunchtime, at special sessions that he sometimes holds for his A.P. students during the summer, and on every weekend in the months leading up to the exam. He has been teaching the class for six years, and all but one of those years, all of his 25 or so students have passed the exam, many with top scores. Two years ago, one of his students made headlines when he received a perfect score on the test, one of only 12 in the world to do so that year.

These success-against-all-odds stories are captivating. It’s hard to overstate how much “Stand and Deliver” — the 1988 movie about an A.P. calculus teacher who overcame the odds when all his low-income Latino students passed the exam — has influenced many advocates’ perceptions about what an A.P. class can do. And things like this do happen; “Stand and Deliver” is based on real events. But they’re anomalous. Yom credits his success to a number of things: a math department that lays out clear expectations from ninth grade on about what students need to know to get to A.P. calculus, a mentor who has taught A.P. calculus at Lincoln High for 16 years and his own ability to devote countless hours to his students. But once Yom is married and has children, he told me, it simply won’t be sustainable to continue spending so much time with his classes.

And stories like Yom’s are rare, in any case. At struggling schools, it’s common that students are strongly encouraged to enroll in A.P. courses, or simply placed in them by their counselors. This difference was apparent to Paige Veliz-Gilbert, a history teacher at Woodson High, when she attended a one-week A.P. training program last year. She is a strong supporter of A.P.s but found that many of the College Board’s trainers didn’t understand the lives of the students she was teaching. Veliz-Gilbert says she remembers asking one workshop leader what percentage of his students complete their homework. “And without batting an eye, he said: ‘All of them.’ He saw the look on my face. I’m lucky to have a 30 percent return on homework.”

Even if students don’t pass the test, there is reason to believe that simply taking A.P. courses is valuable. After all, many students receive passing grades in their courses while still failing the A.P. exam. But because so much focus is on the test — the College Board tracks only participation and outcomes from the tests, not the classes — and because numbers are so much easier to measure than the far more intangible benefits of teaching and learning, the real value of A.P.s can be hard to assess. It seems logical to assume that taking a more rigorous course can have benefits in and of itself: by opening horizons, by sending a message to students that they are capable. And many teachers and students feel that way. Calid Shorter, 17, who was in Fuchs’s A.P. government class this past year, says she was one of his best teachers. “They really care,” he says. “Pushing me into classes has been a benefit — it’s given me more of a go-getter mind-set.”

Frazier O’Leary, who retired this year after teaching A.P. classes at Cardozo high school in Washington for more than 20 years, doesn’t worry about the fact that of his 64 students who took the A.P. literature exam this year, none passed. “I was so excited 11 of them got 2s,” says O’Leary, who also grades A.P. exams for the College Board and teaches A.P. teachers at summer institutes. “A 2 means they understood the text. For some of these kids, it was the first time they were exposed to studying novels or plays. Some have never seen a play. Some don’t own any books at all.” He has no doubt that simply being in his A.P. literature class helps his students succeed in college. “When kids come back after freshman year, they say they got an A or a B in English,” he says — and that they find their college courses easier than his.

But anecdotal accounts like O’Leary’s are not borne out in more systematic studies of the A.P.’s impact. Too often, says Klopfenstein, of the Colorado Evaluation and Action Lab, research confuses correlation with causation; highly motivated students tend to take more A.P. classes, and they also tend to do better in college and graduate on time. But once all the variables, like parental education and income, are stripped away, there is no indication that those who take A.P.s do better in college. “If you don’t control for all the factors, A.P. looks good,” she says. “If you do, A.P. is not so positive.”

The effects of exam scores are clearer. Existing research offers strong evidence that scoring a 3 or more on the A.P. exam predicts greater academic success in college, Klopfenstein says. And while the College Board has published one study indicating that students who get a 2 on the exam may also do slightly better, even Packer at the College Board says that benefit is unproven. “We at the College Board want to be very careful about the language,” he says. “Frequently schools make larger claims for A.P. than we do.”

These findings raise a question: Is it effective to be investing the time and resources in a program whose benefits seem so difficult to pin down? And allocating funds in this way can have perverse consequences. If only a small number of students are truly ready for Advanced Placement, then students who are either unprepared or unmotivated to be admitted have to fill out the class. “Before it was very selective, and now it has gone too far the other way,” says Carlos Veciana, an A.P. teacher in a Miami-Dade charter school. “Now you put 30 kids in a classroom, and 15 have no business being there. And the kids who don’t want to be there, they become disruptive.”

The academic achievement gap, according to most research, usually begins before children even enter school — and research has found that after third grade, on average, these disparities don’t change much. So the only real way to even begin to conquer the inequality apparent in how the A.P.s play out is by addressing the issue much, much earlier. But beyond such a fuzzy and politically distant goal, critics of the A.P. expansion believe there are changes that could be made sooner. Klopfenstein argues that the A.P. program should remain accessible, but that it must be accompanied by regular classes in which students learn skills like note-taking, outlining and intellectual discipline. Others think the mandates on the number of A.P. classes must go, that districts should instead look at which subjects might benefit the most students, rather than arbitrarily drawing a line. Some even advocate for keeping the classes but getting rid of the high-stakes tests at the end.

College Board representatives say they know more help is needed to make the A.P. expansion improve students’ outcomes. At its annual meeting this year, the organization announced that support will soon be available, including a partnership with Khan Academy, an online nonprofit, to offer free test-preparation and course materials for teachers and students in many A.P. subjects. The question is whether directing those resources at passing one specific test makes any sense.

In July, when the A.P. exam results were released, four students out of 162 who took the exams at Woodson passed. “I am resigned to the scores breakdown,” Fuchs wrote to me in an email. “What I will say is I have colleagues at other schools who still get straight 1s, unfortunately. So I know I’m doing something different to get my few kids past the finish line, but I don’t seem to be able to make significant headway with the rest.”

Brian Pick, chief of teaching and learning for D.C. Public Schools, says that A.P.s are only one tool in the district’s overall push to drive up standards. “We all have a deeply held belief that if you set high expectations for kids, they will rise to them, from what course they’re in to what assignment they get,” he says. That turned out to be true in Fuchs’s class. There were 20 1s, four 2s, one 3 and a lone 5: Calid Shorter. Over the course of his junior and senior years, he took five A.P. exams, and A.P. government was the only one he passed. “That’s the only one I truly studied for,” he told me, because it actually interested him. He just started his first semester at Sewanee, the University of the South, near Knoxville, Tenn.

Fuchs started teaching a new crop of seniors last month. She’ll have a chance, once again, to try out the A.P. curriculum on her students. Like any good teacher, Fuchs is an optimist, and she believes that her next group of students will do better. Still, the tone of her email turned philosophical at times. “I think schools embrace A.P. in large part due to the drive for ‘rigor’ and the lack of interest in truly exploring what that means,” she wrote. “It is easier for a district to purchase an outside program and its definition of rigor.”

This story was reported in partnership with The Investigative Fund at The Nation Institute, now known as Type Investigations.