.|  Baltimore Ecosystem Study
Crime and green infrastructure
  • Morgan Grove, Northeastern Research Station, USDA Forest Service
  • Austin Troy, University of Colorado, Denver

This branch of research seeks to answer two questions: 1) how does green infrastructure affect the incidence of crime and; 2) how does crime level condition the way that green infrastructure is perceived?
Crime-mitigation as a benefit of urban trees and landscaping

There has been some debate in the literature over the years about the relationship between crime and vegetation. One perspective in the literature sees vegetation primarily as affording opportunities for concealment and caching. There is evidence that this may be the case for at least certain types of vegetation, particularly when it is low and offers a visual block. On the other hand, a different branch of the literature has found that vegetation, particularly higher canopy trees, can lead to reduced crime through four potential pathways: 1) trees improve cognitive functioning and reduces aggression; 2) trees encourage residents to spend time outdoors, fostering both interaction and "eyes on the street"; 3) more outdoor time fosters social networks, leads to informal surveillance; and 4) well maintained trees serve as "cues to care" or territorial markers, essentially the inverse of the "broken window theory." In other words, vegetation has the ability to both facilitate or mitigate crime, and the question of which side wins this tug-of-war depends large on the levels of concealment afforded by the vegetation and its level of maintenance.
Our research has determined that in Baltimore City and County, tree canopy cover is significant associated with a large reduction in crime. Summarizing data at the block group level, we regressed a crime index (derived from actual geo-coded crime locations and based on street crimes, such as robbery, burglary and shooting crimes) against tree canopy cover and a number of covariates meant to control for confounding (population density, income, percent single family homes, percent rural, race, median year of construction, amount of protected open space, and amount of agricultural land). Regressions (adjusting for spatial autocorrelation) were also run that broke down the tree canopy cover by private land, public land including rights of way (street trees) and public land excluding rights of way.
The regressions, which explained approximately 85% of the variance, indicated that a 10% increase in tree cover is associated with an 11.8% decrease in crime. When tree cover is broken up by private and public land, the magnitude of the effect goes down slightly for each, and the more conservative spatial model yields a big difference appears between the public and private tree effects, with the former being nearly 40% larger. This would suggest that planting trees on public lands might yield somewhat higher crime-reduction benefits than planting on private. When public rights of way are not included in the analysis of public trees in the spatial model, the gap between public and private increases relative to the previous model, making the magnitude of the public trees coefficient nearly double that of private trees, suggesting that trees in non-right of way public lands (e.g. parks, public institution lands, etc.) are the most effective components in terms of reducing crime. Our geographic analysis also found that there are some exceptions to this rule: in a few isolated neighborhoods (Brooklyn Park, Wagners Point, and Dundalk ) more trees equate to more crime. This may be related to concealment. There is a considerable amount of lower, early successional, and apparently unmanaged stands of trees both on small residential lots and on larger private institutional/industrial parcels in some of these neighborhoods, particularly adjacent to warehouses, truck yards, factories, etc.

Another study found that actively landscaped front yards also significantly mitigate crime. We surveyed 1,000 parcels throughout the Gwynns Falls watershed, cataloging indicators of yard management. We then regressed these data against measures of street crime, adjusting for a number of socio-economic and housing factors. In this analysis we found that four things were associated with a drop in crime for that immediate area: the presence of a lawn, the number of big yard trees, the presence of mulch around street plantings and indicators of tree and plant pruning. Litter, on the other hand, was positively associated with crime. Our results suggest that active indicators of yard maintenance may deter crime for many of the same reasons mentioned above with regard to tree canopy: well-maintained yards are a "cue to care," they make an inviting environment that results in more "eyes on the street," and they may indirectly foster informal surveillance networks.
How crime can condition perception of green infrastructure
Urban parks are typically considered to be an amenity that can affect location choice decisions and increase neighborhood desirability, as reflected in property values. However, relatively little research exists on whether the existence of high crime rates may turn green space into from an amenity into a perceived liability. Most previous research has found that parks positively impact property values. Our research was among the first to look at how crime level of a nearby park conditions that property value effect. This is important issue in a city like Baltimore, where high crime rates co-exist with an extensive park system. Using hedonic analysis (where the determinants of property price are statistically disaggregated), this study found that there was a significant interaction effect between park proximity and crime levels in determining the price effect of the park. When crime rate is relatively low, parks have a positive impact on property values. That threshold value is between 406% and 484% of the national crime average (which is still lower

than the average robbery rate for Baltimore, at 475%). Near the threshold, the value of parks becomes ambiguous. As crime rates climb above this threshold, the direction of the relationship switches and parks negatively influence home values. The steepness of this negative relationship increases as the crime rate increases. The result then, is a heterogeneous mix of parks that the housing market perceives as amenities and others it sees as liabilities. High and low crime parks are, interestingly, fairly evenly distributed in space, rather than being clustered.
This finding suggests, then, that parks, usually construed to be an amenity that can serve as an attractant in location choices, may not always be. In fact, an unsafe park is a far bigger deterrent for a potential mover than is a neighborhood with no park at all.
Personnel involved:
Austin Troy, J. Morgan Grove, Jarlath O'Neil-Dunne, Ashley Lidman
A. Troy, J.M. Grove and J. O'Neil-Dunne. 2012. The Relationship between Tree Canopy and Crime Rates across an Urban-Rural Gradient in the Greater Baltimore Region, Landscape and Urban Planning. 106: 262-270.
A. Troy and J.M. Grove. 2008. Property Values, parks, and crime: a hedonic analysis in Baltimore, MD. Landscape and Urban Planning. 87:233-245.
Lidman, Ashley (2008) Vegetation, neighborhood satisfaction, and crime : case studies in Baltimore, MD [MS] Rubenstein School of Environment and Natural Resources
A. Troy. June 2012. The relationship between trees and crime across an urban-rural gradient in greater Baltimore. Alliance for Community Trees. Invited webinar presentation.
Troy and M. Grove. October 2010. Relationships between Vegetation and Crime in the Baltimore Metro Region. Baltimore Ecosystem Study Annual Meeting, Baltimore County, MD
Troy and M. Grove. June 2008. The interactive effect of parks and crime on property values in Baltimore, MD. International Symposium on Society and Resource Management, Burlington, VT.
Troy and M Grove. October 2007. The interactive effect of parks and crime on property values in Baltimore, MD. Baltimore Ecosystem Study Annual Meeting, Baltimore, MD.
Press coverage
Baltimore Sun: http://articles.baltimoresun.com/2012-05-19/news/bs-gr-trees-crime-20120518_1_mature-trees-fewer-break-ins-criminals
Boston Globe: http://www.boston.com/lifestyle/green/greenblog/2012/10/trees_may_deter_crime_in_urban.html
Grist: http://grist.org/cities/in-baltimore-the-gods-will-not-save-you-but-the-trees-will/
The Atlantic Cities (Part of Atlantic Monthly): http://www.theatlanticcities.com/neighborhoods/2012/05/can-trees-actually-deter-crime/2107/
Planetizen: http://www.planetizen.com/node/56948
Vancouver Sun: http://www2.canada.com/vancouversun/news/archives/story.html?id=5bb5bac7-65fc-4670-a75c-5abf82bed2c1
Calgary Herald: http://www.calgaryherald.com/technology/Greener+cities+safer+cities+with+less+crime+researchers/6709877/story.html