A brand-new multidimensional approach to student-school matching

I’ve always had a sweet spot for education. Learning new things is fun and exciting, and there is so much to learn. One of such things I stumbled across recently was school choice, and, in particular, models to match students with schools in the most efficient and socially-optimal manner. Picture this: thousands of stressed out parents trying to ensure their kid is accepted into a nice high school in New York City. Not optimal.

Interestingly, the top-of-the-line economic models that govern today’s school choice in New York City, Boston and even Minnesota also govern assignments of doctors to hospitals, donor kidneys to patients and many other real-world problems. One of these problems is financial markets: a vast, highly regulated field that is prone to manipulation. Having spent a considerable time studying the markets, I was instantly fascinated by the matching models. And then I had to reengineer them.

One of the core problems I saw with the current matching models is a single-scale evaluation of matching success. For example, in NYC, all students are asked to rank their destination schools on a linear scale. As anyone in any negotiation and bargaining application will tell you, the single scale or single-issue bargaining is guaranteed to lead to a disaster. There are no happy outcomes in that dog-eat-dog I-win-you-lose world.

Instead, what successful systems, those that make people happy, have is a multidimensional approach: a quality that may be important to one person can be completely irrelevant to another, and a natural trade follows that increases aggregate happiness.

The outcome is, on paper at least, makes a lot more sense to me. But don’t trust my opinion, see for yourself. And please let me know what you think.