Report Study shows urban vegetation correlates with reduced crime rates.

One school of thought has it that inner-city trees and shrubs make convenient hiding places and covered escape routes for criminals. Another, supported by an increasing body of evidence, argues that urban foliage may actually reduce crime. Where previous studies have tended to focus on individual housing blocks or, at best, neighborhoods, new research out of Temple University is among the first to examine the issue at the city-scale. TU researchers analyzed the relationship between vegetation concentration and crime for the whole of Philadelphia.

The researchers broke the city down into 363 "tracts" identified from socioeconomic census data, each containing between 100 and 8,000 people. The data, taken from the years 2005 to 2009, was also used to assess poverty and education levels in these tracts. Vegetation coverage was assessed from satellite imagery from 2005, courtesy of NASA's Landsat 7. Recorded incidents of aggravated assault, robbery, burglary, and theft were drawn from the University of Pennsylvania's Neighborhood Information System CrimeBase, also for the year 2005.

Theft, just to be clear, is the illegal act of taking something without permission. Robbery is the same, but with violence (or the threat of it). Burglary is breaking into a building in order to a commit crime (often, but not necessarily, theft or robbery.) The researchers ignored incidents of murder and rape on the grounds that their relative rarity renders the comparison senseless at tract level.

The team undertook a series of statistical analyses of the data, including multivariate ordinary least squares regression. This allowed them to test the "explanatory power" of foliage on crime while controlling for other influences on it, like population density, poverty, and education. Other techniques allowed for the control of the spatial contagion of crime.

Read more here.

The correlation is related to funding that part of the city gets. Obviously the areas with higher crime rates likely are more poverty stricken districts of the city, and results of that are under or unmaintaned parks and gardens and streetside foliage, in addition to the police forces likely being primarily given routes in the richer neighborhoods. Planting more trees won't change any of that; it will just make the poverty areas look better, and as a result it's possible that there is a migration of richer people to those areas than was previously seen (directly linked to upgraded aesthetic of the area), which would likely snowball to mean more police are patrolling those areas, reducing the crime rate locally, but most likely what will happen is it just shifts to a different area or congregates with other high crime regions.

Nor did it mention the historical demographics of the districts or overall aesthetic appeal. Nice families tend to move to nicer neighborhoods; nice poor families tend to move to nicer poor neighborhoods. There are other significant variables than a few core demographics.