Vacant and abandoned buildings are a long-standing problem for older industrial cities. Many of these cities have experienced sustained population loss and must manage an oversupply of vacant houses at levels that have not substantially changed since the Great Recession (Harrison and Immergluck, 2021). Vacant houses are often disproportionately located in poor neighborhoods and exert a fiscal toll on local governments through foregone tax revenue and need for maintenance (Harrison and Immergluck, 2021; Mallach, 2018). For example, Baltimore, Maryland is conservatively estimated to have >15,000 vacant buildings (City of Baltimore, 2021a), most of which are concentrated in Black communities with histories of discriminatory housing policies.

Prior research suggests that vacant properties negatively impact public safety and health. For example, in Philadelphia, vacant buildings and lots have been associated with increased rates of aggravated assault (Branas et al., 2013). Residents in that city also reported that vacant and abandoned houses increase injury risk, attract illegal dumping and rodents, contribute to higher levels of anxiety and fear, and undermine the cohesion and overall well-being of the community (Garvin et al., 2013). Other studies have found associations between vacant buildings and increased risk of fire in occupied dwellings, deaths related to drug dependence, rates of sexually transmitted disease, and all-cause premature mortality (Schachterle et al., 2012; Cohen et al., 2003; Hannon and Cuddy, 2006). In one study, higher densities of housing code violations were associated with increased emergency department use for children with asthma (Beck et al., 2014).

The connection between vacant buildings and these negative safety and health outcomes may be explained through multiple, interrelated pathways. One set of theories focuses on the role of visible physical disorder and the risk of crime. For example, broken windows theory posits that dilapidated vacant buildings signal to passers-by that an area is not cared for or monitored, and that crime or violence is tolerated (Kelling and Wilson, 1982). Similarly, sites of physical disorder are theorized as places of convergence between people like to commit a crime, potential targets, and the lack of guardians willing to intervene (Cohen and Felson, 1979). Critiques of this framework have emphasized the role of poverty and concentrated disadvantage, more than physical disorder alone, as drivers of crime, fear of crime, and ill-health (Taylor, 2018). From a social ecological perspective, vacant buildings foster an environment that erodes social interactions and collective efficacy, which may, in turn, harm health outcomes (Taylor, 2018; Sampson et al., 1997). In addition, vacant and abandoned homes may promote chronic stress and behavior changes, such as decreased physical activity, and harm health through toxic exposures (Garvin et al., 2013).

Amid the theoretical debate, several U.S. cities have responded to the problems posed by vacant buildings by expanding demolition programs (Mallach, 2018). However, there is mixed evidence for whether the demolition of vacant buildings by itself improves neighborhood safety and health. Prior research tends to focus on crime as the primary outcome. In Detroit, which undertook one of largest scale programs after the Great Recession, demolitions were associated with an 11% reduction in firearm assaults (Jay et al., 2019). Studies in similar urban contexts found slightly smaller crime reductions that dissipated a few months after the demolition (Stacy, 2017), decreased sharply beyond the parcel level (Wheeler et al., 2018), or were only significant for burglary and theft (Spader et al., 2016). In one recent study, crime hotspots shifted away from demolition activity while citywide rates remained unchanged (Frazier et al., 2013). Elsewhere researchers found no association between demolitions and most types of crime (Han and Helm, 2020; Kim and Wo, 2020). Most of the research examined either the time period immediately before and after demolition or over the longer term (beyond one year) but did not compare effects over varying time periods.

By comparison, relatively little research measures the impact of vacant housing demolition on health outcomes other than crime. Most studies instead focus on the role of cleaning and greening vacant lots, which is hypothesized to improve health by reducing the risk of violence, lowering stress levels, and increasing opportunities for physical activity (Branas et al., 2016; Sivak et al., 2021). Yet the unique contribution of vacant housing demolition to these pathways remains uncertain. In one study, researchers found that self-reported measures of physical and mental health were not significantly different for residents living in neighborhoods undergoing demolition as part of an urban renewal program, compared to residents in control neighborhoods (Egan et al., 2013). Otherwise, studies on the health effects of demolition tend to focus instead on specific harmful exposures such as lead dust, which may worsen air quality and exacerbate asthma in the short-term (Dorevitch et al., 2006; Jacobs et al., 2013). Given this evidence and prior theory, we posit that vacant housing demolitions may be associated with changes in health through three main pathways: reduced stress, reduced risk of injury, and fewer hazardous environmental exposures. We theorize that one or more of these pathways could reasonably manifest in the change in overall emergency department (ED) visits from neighborhoods where vacant homes are being demolished. ED visits represents a costly form of healthcare use that is sensitive to neighborhood racial composition and housing tenure (Li et al., 2003), though previous studies have not, to our knowledge, examined its relation to housing demolitions. In addition, given its time-sensitive nature, ED visits may be particularly responsive to demolitions and may differ for adults and children due to underlying vulnerabilities and critical periods.

Baltimore City provides an important test case for examining the relationship between demolition of vacant buildings, crime, and health. Like other cities that have struggled with population loss, such as Buffalo, NY and Detroit, MI, Baltimore has directed public funds to demolition to reduce its oversupply of vacant housing. Over the past decade, Baltimore and the state of Maryland have demolished >3500 vacant buildings, primarily attached single- and multi-family rowhouses, through the Vacants to Value program and the Project CORE program (Baltimore City Bureau of the Budget and Management Research, 2013; Maryland Department of Housing and Community Development, 2016). Vacants to Value is a city-led neighborhood revitalization initiative which in 2013 committed $100 million over ten years to housing code enforcement and demolition (Baltimore City Bureau of the Budget and Management Research, 2013). Project CORE, which the state of Maryland launched in 2016 after the in-custody death of Freddie Gray, added $75 million into neighborhood revitalization in Baltimore, with an emphasis on demolishing vacant housing and reducing crime (Maryland Department of Housing and Community Development, 2016). Concurrent with these demolitions, several other initiatives, including the 2018 Violence Reduction Initiative (VRI), were implemented to concentrate city services in parts of the city with elevated gun violence (Duncan, 2018).

With its large investment in demolitions, we sought to examine the association between demolishing vacant housing in Baltimore City with changes in crime and ED visits, examining outcomes both immediately following a demolition and over longer periods of time.