Work

Can a Twitter-Based Measure of Racial Heterogeneity Explain Neighborhood Crime Differences? Proposing an Ambient Population Measure of Social Disorganization

Year: 2026
Type: article
Abstract: Social disorganization theory traditionally studies neighborhood social dynamics through residential demographic indicators. However, insights from routine activities theory show that people spend time outside their home neighborhoods, suggesting that the populations who use places may differ from those who reside there. Building on this tension, this study uses an ethically informed approach to introduce a social media-derived measure of ambient racial heterogeneity and assess whether dynamic population compositions explain crime variation across neighborhoods. Using a Gaussian Naïve Bayes approach to estimate race with geotagged mobility data, we track the spatial footprints of 6,547 Twitter users who posted in Boston in 2018. These data are used to construct census-tract measures capturing the racial diversity of neighborhood users rather than residents. Validation indicates 90.5% agreement with reference classifications and demographic comparability to Boston’s population. Negative binomial models show that ambient racial heterogeneity is positively associated with violent and nonviolent crime in nonresidential tracts.
Source: Figshare
Language: English
Open Access status: green
OpenAlex ID: W7226002979
FWCI:
Cites:
Cited by:
Related to: 10