A Trajectory Analysis of Residential Burglaries Known to the Police in the Central 23 Wards of Tokyo (Symposium: The Trend of Crime and Delinquency-Changes in Quantity and Quality in Contemporary Japan-)

Bibliographic Information

Other Title
  • 東京23区における空巣の認知件数の軌跡分析 (課題研究 犯罪・非行の量的・質的変化動向)
  • 東京23区における空巣の認知件数の軌跡分析--セミパラメトリック混合ポアソンモデルによる検討
  • トウキョウ 23ク ニ オケル アキス ノ ニンチ ケンスウ ノ キセキ ブンセキ セミパラメトリック コンゴウ ポアソン モデル ニ ヨル ケントウ
  • セミパラメトリック混合ポアソンモデルによる検討
  • An Examination with Semiparametric Mixed Poisson Models

Search this article


This study examines trajectories of the geographic distribution of residential burglaries in the central 23 Wards of Tokyo over 10-year period from 1996 to 2005, using semiparametric mixed Poisson models. The data consist of official records of criminal incidents of residential burglaries while residents were out ("AKISU") known to the Tokyo Metropolitan Police Department. These incident records were address-geocoded and mapped onto a digital map and then aggregated into a offense-count dataset by each of the 3,133 CHO-CHO-MOKUs. A SAS procedure for trajectory analysis, Proc Traj, was applied to this dataset, assuming Zero-Inflated Poisson distributions of the dependent variables and quadratic forms of the baseline trajectories. The main findings are as follows : 1. Sixteen groups of trajectories are identified, based on stepwise comparisons of model fitness using the BIC (Bayesian Information Criteria). 2. A majority of these 16 trajectory groups show reverse-U-shaped curves with peaks around the year 2001. However, a few trajectory groups show different shapes, such as slowly increasing or decreasing trajectories over the 1O-year period. 3. Three trajectory groups show particularly high peak levels of offending and strong reverse-U shapes of trajectories. It appears that these trajectories reflect the sharp increase and decrease of offenses with a new type of modus operandi called lock pickings, as well as the impact of crime prevention efforts that took place in these areas in response to the increasing. crime rates. 4. Two of the 16 trajectory groups show slowly increasing offense trajectories over the 10-year period. The CHO-CHO-MOKUs with these trajectories appear to be spatially clustered with one another. These results suggest that future offense trends in these areas should be monitored closely for possible emergence of high crime-risk areas.


Citations (1)*help

See more

Details 詳細情報について

Report a problem

Back to top