Crime on the mass transit system in Hong Kong: a hotspots and harmspots trajectory approach

Yiu Ming Ng, Barak Ariel*, Vincent Harinam

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Purpose: A growing body of literature focuses on crime hotspots; however, less is known about the spatial distribution of crime at mass transit systems, and even less is known about trajectory patterns of hotspots in non-English-speaking countries. Design/methodology/approach: The spatiotemporal behaviour of 1,494 crimes reported to the Hong Kong’s Railway Police District across a two-year period was examined in this study. Crime harm weights were then applied to offences to estimate the distribution of crime severity across the transit system. Descriptive statistics are used to understand the temporal and spatial trends, and k-means longitudinal clustering are used to examine the developmental trajectories of crime in train stations over time. Findings: Analyses suggest that 15.2% and 8.8% of stations accounted for 50% of all counted crime and crime harm scores, respectively, indicating the predictability of crime and harm to occur at certain stations but not others. Offending persists consistently, with low, moderate and high counts and harm stations remaining the same over time. Research limitations/implications: These findings suggest that more localised crime control initiatives are required to target crime effectively. Originality/value: This is one of the only studies focusing on hotspots and harmspots in the mass transit system.

Original languageAmerican English
Pages (from-to)908-921
Number of pages14
Issue number5-6
StatePublished - 7 Nov 2023

Bibliographical note

Publisher Copyright:
© 2023, Emerald Publishing Limited.


  • Harmspots
  • Hong Kong
  • Hotspots
  • Mass transit system
  • Trajectory analysis


Dive into the research topics of 'Crime on the mass transit system in Hong Kong: a hotspots and harmspots trajectory approach'. Together they form a unique fingerprint.

Cite this