Identifying Upzoning Opportunities

urban
prediction
machine learning
Published

December 22, 2023

For a “Public Policy Analytics” class at Penn, Laura Frances and I analyzed conflicts between anticipated and current zoning. Using a random forest model trained on historic Philadelphia construction permits, we identified areas of the city with restrictive zoning that would potentially hinder anticipated residential development. We then considered how the model could be used to inform efforts at zoning reform and more strategic planning meant to alleviate issues of housing affordability across Philadelphia.