When charter schools began a quarter-century ago, one of the many reasons critics said school choice wouldn’t work is that low-income parents couldn’t possibly acquire enough information to choose a school. They wouldn’t know how to conduct a search. They wouldn’t have enough patience to fill out all the forms necessary. Smarter parents, expert at manipulating bureaucracies, would have all the advantages. The inevitable result: the upper-tier schools would “cream” the best students and the poor would be stuck with failing inner-city schools.
But the critics who predicted this gloomy future did so before the World Wide Web was created. In a piece for the Washington Post Magazine, Thomas Toch, who heads FutureEd, describes a new tool that makes matching students with schools much more efficient—thanks to a discovery that earned its creator a Nobel Prize in economics.
The District of Columbia has one of the nation’s more robust education markets, with about half of the students in secondary schools attending charters and half going to more traditional public schools. Although the district guarantees a spot in a neighborhood school for every student, only 27 percent of children go to their local school. The remainder go elsewhere in the city, thanks to an app called My School DC.
When charter schools began in Washington in the 1990s, each school had their own application system. Parents who wanted to choose schools had to apply, on paper, to all the schools they thought their children would like to attend and had to wait for the envelopes that would reveal whether a student was accepted or rejected. Parents had very limited information about what the schools offered.
This system, Toch writes, led to “oversubscribed schools pulling names out of paper bags, families pitching tents on sidewalks—or paying others to camp out for them—to get to the front of wait-list lines, and schools cherry-picking applicants to get the most attractive students.” This led to “a system favoring the wealthy, the well-educated, and the well-connected.” Furthermore, many parents didn’t pick a school until September or even October, leaving schools with poor information about how many students they would have in a school year.
With My School DC, which began operations in the 2014-15 school year, the process is much fairer. Parents get information from school websites, open houses, and the annual Ed Fest, where schools set up booths and talk to parents. Every parent then sets up an account at My School DC and gets a lottery number generated at random. Students who don’t get into their top-ranked schools are put on the waitlist for schools they’ve given a lower rank. Families can see where they are on a particular school’s wait list.
The result is a much fairer system. Of course, not all students get to go to the best schools, but My School DC enables more students to go to schools they prefer. In addition, because everyone is part of one system controlled by an algorithm, wealthy parents can’t have advantages because they can “stand in lines, lobby principals, or complete scores of applications.” Schools save money because they don’t have to hire people to process application forms.
In addition, My School DC gives administrators information that didn’t exist about what sorts of programs parents in particular areas want, so they can add popular programs (music, foreign languages) in areas that didn’t have them.
The system is the result of research done by Stanford economist Alvin Roth. In 1995 Roth, who was teaching at Harvard at the time, devised an algorithm that made it much more efficient for medical school graduates to find hospitals for their residencies. He’s since done research on creating markets for placing law school students in clerkships, conducting Internet auctions, and he’s currently working on kidney donations.
In 2003 Roth, graduate student Parag Pathak (who’s now an economist at MIT), and economist Atila Abdulkadiroglu (who’s now teaching at Duke), were asked to fix the system with which 80,000 eighth graders in New York City picked their high schools. The existing system was “surprisingly stressful” since many parents, instead of choosing the schools they wanted, picked ones they thought were less popular in the hopes of getting in. The system was rife with cronyism and at the end of the process over a third of the students still didn’t know where they’d head for ninth grade.
Roth’s system creates a lot of temporary matches for the second, third, and fourth choices students make, calculating and recalculating to ensure that every student gets as high a choice as possible. Not everyone gets their first choice, but far more students get their second choices than they would have before My School DC was created.
The result, Roth says, is that the school-selecting algorithm puts “more students in schools they want to be in. Children who grow up in poorer neighborhoods shouldn’t be condemned to go to poorer schools.”
For his discoveries in enhancing market efficiency, Alvin Roth won the 2012 Nobel Prize in Economics.
My School DC does not solve every education problem. Public schools in the District of Columbia are about 15 percent white, and most white students live in wealthy Ward 3. A feature of My School DC called “in-boundary preference” ensures that families who choose their neighborhood school have preference over students from other parts of the district. There are no charter schools in Ward 3, and most parents living in that ward choose their neighborhood school. So My School DC doesn’t do much to deal with segregation.
But My School DC does make the competition between charter schools and traditional public schools in the District of Columbia stronger, fairer, and more robust. The result is that both charter schools and public schools in the district are better than they were before Alvin Roth invented his market-matching algorithm.