New York City’s public schools have struggled with overcrowding for decades—and in recent years, the problem has worsened. In the 2012-2013 school year, 43.5 percent of public school students went to school in overcrowded buildings. For these kids, that means fewer teachers, larger class sizes, and fewer resources.
But overcrowding turns out to be an extremely complicated problem to address, because enrollments ebb and rise annually and the demand for space in different schools isn’t uniform: Some have empty seats, while others are bursting at the seams. And the next year, the situation could be reversed. The city tackles the problem on many fronts. On one hand, it tries to create room for kids in currently overcrowded schools by rezoning, which can become incredibly contentious. (New York Times Magazine reporter Nikole Hannah-Jones recently wrote an extensive story on her daughter’s school in Brooklyn became a battleground for parents embroiled on both sides of the rezoning war.) Another longer-term fix: constructing new school buildings to accommodate higher demand, which the Department of Education has planned to do in their 2015-2019 vision.
But in order to better plan for geographical fluctuations in demand, the DOE tries to peer into the future, projecting which areas will have growing school enrollment rates in a few years. Now it’s putting out a call to all concerned data wonks to help make these future predictions sharper. “As the largest school district in the United States, we are always looking to improve data models,” Elizabeth Rose, deputy chancellor of New York City’s Department of Education, said in a statement to CityLab. “Changes in data technology may enable better projections on a more local basis.”
The DOE has announced a competition in which they’re asking people to submit more nimble, more detailed methods of reckoning future enrollment. Specifically:
The City is looking for alternative data models that can enable more granular and precise student projections—ideally, detailing the number of prospective students at different age groups by city block with regularly updated projections at 1-year, 2-year and 5-year increments.
Interested parties have three weeks to sign up, after which they’ll be given access to a portion of DOE’s enrollment dataset. If their predictive model makes the next cut, they get to take their ideas deeper, using additional datasets and $5,000. There’s another $15,000 in the offing if their model wins in March 2017. Of course, this winning model will be used by the city to make future predictions. And if other cities are interested in replicating it, the DOE says it would share the idea.
This is the second time the city has issued a call for innovation in an effort to get ahead of an issue, and this one is particularly significant. “There are overcrowded districts in every borough and every zone,” says Leonie Haimson, founder and executive director of Class Size Matters, a nonprofit that advocates for smaller class sizes. School district 24 in Queens and 20 in Brooklyn*, for example, are among the most crowded in the city. The conditions there help seal the achievement gap between the predominantly immigrant, low-income, and minority students, and their whiter, more well-off peers in less-crowded schools. “In too many schools, especially, in elementary schools in New York City, they’ve lost their art rooms, their science rooms, their music rooms, sometimes their libraries and gyms, because they’ve gotten so overcrowded that they really need that space for general instruction,” Haimson says. “It’s also a huge issue for special-needs kids who have seen their intervention rooms lost to overcrowding, and are often getting their services in hallways, closets, lunchrooms and stairwells.”
Haimson is among the critics of the New York City’s current method of projecting enrollment. Her own organization’s calculations have shown a greater need for public school seats than the city’s account in the past. The biggest problem, in her mind, is that the current prediction model underestimates how new housing will drive enrollment in nearby schools. (The DOE maintains that their current model is reliable and frequently updated.)
In their new call for innovation, the DOE encourages prospective participants to use real estate and housing data, as well as consumer behavior data, information on day-care services, pedestrian and transit data, and of course, demographic trends. The winning model, whatever it turns out to be, could potentially serve both the goals of the DOE and the advocates on the ground—and hopefully help ease tensions over a hot-button issue.
Sign up to participate in the challenge here.
*CORRECTION: This post has been updated to reflect the true location of school district 20. It is in Brooklyn, not in the Bronx as previously stated.