Crime Reduction through the Removal of Vacant Buildings: A Baltimore City Study
The City of Baltimore has grappled with the challenges of vacant buildings and high crime rates for decades. In this study, scientists sought to understand how the removal of vacant buildings might affect crime, especially violent crime.

Vacant and abandoned buildings are a common feature in many post-industrial U.S. cities. They symbolize urban blight and are an indicator of disinvestment and neglect. Large numbers of vacant structures in an area provide locations for criminal activities, are consistent predictors of violence, and reduce residents’ quality of life. The objective of this research was to examine the effect of vacant building removal on violent and property crimes in Baltimore between 2014 and 2019.
Key Findings
- Vacant buildings have a negative impact on neighborhoods and the quality of life for residents. They are consistent predictors of crime in urban areas.
- Building removal led to a decrease in overall crime, particularly violent crime. This suggests that vacant building removal may be a viable crime reduction strategy.

Approach
Northern Research Station scientists and collaborators obtained data on vacant building deconstructions and demolitions between 2014 and 2019 from the Baltimore City Department of Housing and Community Development. The Department also provided a database of vacant buildings during the study period which were used as control locations. Scientists obtained crime reports for the same time span. Analysis found that vacant building removals in Baltimore were associated with reduced overall crime, and particularly violent crime, in the same neighborhoods.
Outcomes
This study suggests that ‘right-sizing’ shrinking cities by removing vacant buildings may lead to a reduction in overall crime including violent crime.
Building removals may also provide other benefits such as preventing building collapses, eliminating blight, and stabilizing housing markets.
Key Personnel
Collaborators
Michelle Kondo, USDA Forest Service Northern Research Station
Rebecca Fix – Johns Hopkins University
Arianna Gobaud – Columbia University
Christopher Morrison – Columbia University
Jon Jay – Boston University