You are here

Battling urban blight with big data

Hopkins scholars and Baltimore leaders are teaming up to use big data to understand the dynamics of vacant housing in the city — and help ensure resources are used effectively
Posted May 11, 2016
Michael Braverman, Tamás Budavári, and Philip Garboden Michael Braverman, Tamás Budavári, and Philip Garboden

What blighted blocks in Baltimore are more likely to need demolition — and which are not? When are strategic interventions from city agencies likely to push rehabilitation — not just of a single building, but of an entire neighborhood? Tamás Budavári and Michael Braverman don't know for sure. But thanks to support from a 21st Century Cities Initiative seed grant, they're getting closer to finding out.

Budavári, an assistant professor in the Whiting School's Department of Applied Mathematics and Statistics, and Braverman, a deputy commissioner for Baltimore Housing and a 1981 alumnus of the Krieger School, began conversations a couple of years ago about building a unique database and accompanying statistical tools that can better identify vacant and abandoned properties throughout Baltimore. Once built, the tools will inform policies and interventions the city implements to revitalize or remove blighted buildings. Eventually, the team hopes to produce a model to share with other cities, enabling them to make immediate and long-term changes to reinvigorate struggling urban areas.

"Sometimes I joke around that we're creating glasses that will allow us to see new dimensions of the vacant-building universe," Braverman says. "But that’s what we're doing — this will deepen the understanding that informs our daily decisions."

For example, his current data — officially filed housing violations, construction permits, court proceedings, and tax delinquency — tells him that Baltimore's vacant, uninhabitable properties number about 17,000. But the full number of long-term unoccupied properties — those still in decent shape, but not lived in — is much larger. The products of foreclosure, unprobated estates, and other mechanisms, these properties eventually become uninhabitable, dragging their neighborhoods down with them. Unoccupancy is not directly measured, so information vital to making decisions about demolition and other revitalization methods is missing.

Yet Budavári, who holds a joint appointment in the Krieger School's Henry A. Rowland Department of Physics and Astronomy, and Philip Garboden, a Krieger School doctoral student and a member of the Poverty and Inequality Research Lab, are working with Braverman's team to develop a proxy measure to determine a property's occupancy status. Specifically, they're merging the city’s administrative housing data with parcel-level data, such as utility and water usage, and whether the postal service considers an address deliverable. Braverman will use that new statistically derived measure to inform the best set of interventions for a blighted neighborhood.

Although Budavári, Garboden and Braverman haven't yet determined the perfect inference method for occupancy, they've made significant strides toward that, and other, discoveries. Seed grant funding from the Institute of Data-Intensive Engineering and Science has enabled the team to achieve a major milestone: building a customized database system.

"If every research question is slow or expensive to investigate, then this project won't work," Budavári says. "Using our data solution and analysis methods, we'll be able to quickly and rigorously answer not only the questions we have now, but also those we want to ask later in light of new data sets."

New seed funding from the 21st Century Cities Initiative, a priority of the Rising to the Challenge campaign, will help the team build on these early efforts and take the project's next steps. It also enables Budavári and Garboden to bring more graduate and undergraduate students into the project — creating important opportunities for training young researchers.

"Without seed grants, it's hard to get money for this kind of project," Garboden notes. "This will allow us to be quick, responsive, and creative."

The creativity will come in as Garboden, Braverman, and other team members develop broader, long-term research questions that Budavári can help answer using statistical inference. Questions such as, "Where in the city is abandonment increasing — and why?" Or, "What is the tipping point in the level of vacancy after which rehabilitation becomes unlikely and demolition may be necessary?" And eventually, perhaps: "Can we predict whether a property will become vacant?"  The answers to these questions could inform preventative measures that have the potential to save cities millions in demolition costs.

For Braverman, the project's potential carries the promise of both professional — and personal — satisfaction.

"My life has been shaped by my experience at Johns Hopkins," he says. "I feel a great kinship to this town. It's my adopted city, and Hopkins was the vehicle for that adoption. I'm excited to continue that relationship in my professional life."

Make a gift using our secure online giving form

To learn more, please contact:
Andrew Rentschler
Executive Director
Rising to the Challenge campaign
Phone: 410-516-0470