Exploring the Impact of Ambient Population Measures on London Crime Hotspots

• In Event: Crime and Place: Shifting Crime Risks By Night and Day

Abstract

Crime analysts need accurate population-at-risk measures to quantify crime rates. The number of residents in an area is commonly used. However, for some types of crime the denominator should capture the ambient population. The aim of this research is to identify, from a suite of available population measures, the most reliable population-at-risk measure for the crime of ‘theft from the person’ in London, UK. We compare the 2011 census workday population, population 24/7 estimates, geo-located twitter messages, and aggregate mobile telecommunications data. We correlate the population measures against crime volumes to identify the strongest predictor, use the GI* statistic to identify statistically significant clusters of crime under alternative denominators, and explore the locations of clusters, comparing those that are significant under the ambient population measure to the residential population. We find some areas that exhibit statistically significant hotspots using both the ambient and residential crime rates. This hints at an environmental backcloth that might be indicative of a mix of both crime generators and attractors. We also find that some regions emerge that are largely residential and yet only exhibit hotspots under the ambient population. These might be places with a higher proportion of crime attractors to stimulate crime, but fewer generators to attract the large volumes of people.

Authors

• Nicolas Malleson, University of Leeds

• Martin Andresen, Simon Fraser University