Even the Best “Game Theory” Algorithm Cannot Ensure Justice via School Choice

The NY Times just explained the sophisticated school-choice matching system that places students in New York City’s high schools.  The city’s  formula models econometric “game theory” to provide, it is said, optimal choices both for the students submitting applications to the city’s high schools and for the schools seeking the students they want:

“Students list their favorite schools, in order of preference (they can now list up to 12).  The algorithm allows students to ‘propose’ to their favorite school, which accepts or rejects the proposal.  In the case of rejection, the algorithm looks to make a match with a student’s second-choice school, and so on… In 2004, the first year that students were sorted in this way, the number who went unmatched plummeted, from 31,000 in 2003 to about 3,000—still a lot of disappointed teenagers.  That year, and every year since, the algorithm has assigned roughly half of all students to their first-choice schools; another third or so have been assigned to their second or third choices.”

The New School Center for New York City Affairs explains in a 2009 report how high school choice came to replace assignment of high school students to their neighborhood high schools: “New York City has long had the largest system of school choice in the country, and (Joel) Klein sought to expand it further.  He eliminated zoned neighborhood high schools in large swaths of the Bronx, Manhattan and Brooklyn and required all eighth graders to fill out applications to high schools.”

Research demonstrates that school choice works best for children whose parents are able to research and play the system.  In the same report, the Center for New York City Affairs concludes: “The system of school choice assumes each child has a parent or other adult who is willing and able to take the time to tour schools and fill out applications.  In fact, many children have no such help… Many middle-school guidance counselors, charged with helping students fill out their high school applications are overwhelmed by huge caseloads and the sheer complexity of giving meaningful advice about 400 different schools.  The weakness in school guidance is particularly an issue for students who don’t have well-educated, English-speaking parents to help them navigate high school admission.”

In New York City, a significant number of students, many immigrants or students with great needs, fail to participate in the school choice program, despite that no child these days has a default high school to attend.  These are the students New York City designates as “over the counter” students.   A scathing 2013 report from the Annenberg Institute for School Reform at Brown University explains, “Every year, some 36,000 students who enroll in New York City high schools without participating in the high school choice process are labeled as ‘over-the-counter’ or OTC students and are assigned a school by the New York City Department of Education.  These young people are among the school system’s highest-needs students—new immigrants, special needs students, previously incarcerated teens, poor or transient or homeless youths, students over age for grade, and students with histories of behavioral incidents in their previous schools.”  The Annenberg researchers conclude: “OTC students are disproportionately assigned to struggling high schools… OTC students are disproportionately assigned to high schools that are subsequently targeted for closure or that are undergoing the closure process.”  “We found twenty-eight high schools that consistently fell within the highest 20th percentile of OTC rates each year from 2008-2011.  These schools’ average OTC rate was 29 percent…. Moreover… these twenty-eight high schools had significantly lower average eighth-grade proficiency scores, college readiness indicators, and graduation rates, as well as significantly higher dropout rates.”

The recent NY Times article concludes that despite the operation of a game-theory based algorithm, the problem continues: “Good schools remain a scarce resource, especially in poor neighborhoods, and low-income and low-performing children are still more likely to end up in underfunded schools. Sean Corcoran, associate professor of educational economics at New York University’s Steinhardt School of Culture, Education and Human Development, has studied the choices made by low-achieving students, who are disproportionately poor.  He found that the algorithm matches low-and high-achieving applicants with their first-choice schools at roughly the same rate.  But… ‘Lower-achieving kids are applying to lower-achieving schools and ranking them as their top choices.’  It seems that most students prefer to go to school close to home, and if nearby schools are underperforming, students will choose them nevertheless.”

Of course there is another strategy—one that is now being promoted by Mayor Bill de Blasio’s administration: improve rather than close struggling schools; strive for quality at every school.  Last week, according to the Brooklyn Daily Eagle, schools chancellor Carmen Farina announced that 45 New York City schools have now been paired with community agencies to become full-service, wrap-around Community Schools under a $52 million, four-year Attendance Improvement and Dropout Intervention grant in partnership with the United Way of New York City.  Universal school improvement will be challenging for the de Blasio administration, especially after the years of growing inequity under Mayor Bloomberg’s administration.

School district strategies based on universal school improvement and on choice and competition represent two very different values.  One seeks the larger goal of serving all children; the other is directed at a select group of children who are thought most likely to succeed.  My own preference is reflected in the statement from the late Senator Paul Wellstone that appears every day at the top of this blog:  “That all citizens will be given an equal start through a sound education is one of the most basic, promised rights of our democracy.  Our chronic refusal as a nation to guarantee that right for all children…. is rooted in a kind of moral blindness, or at least a failure of moral imagination….  It is a failure which threatens our future as a nation of citizens called to a common purpose… tied to one another by a common bond.”

2 thoughts on “Even the Best “Game Theory” Algorithm Cannot Ensure Justice via School Choice

  1. Reblogged this on aureliomontemayor and commented:
    School district strategies based on universal school improvement and on choice and competition represent two very different values. One seeks the larger goal of serving all children; the other is directed at a select group of children who are thought most likely to succeed.

  2. Helpful post, Jan. Thank you. The degrading label, “over the counter” students, reflects both attitude and practice in the system. I sent your post to an economic anthropologist friend who has written about choice, including game theory, and he replied on the run that certainly the more advantaged could play such a system to their advantage and that while it may not be so bad as other systems, it has very little to do with social justice.

Leave a comment