Abstract:

The majority of empirical studies related to crime and security risk for mass gatherings at stadia have focused on general space-time pattern analysis, but in short of considering how various potential crime generators and attractors are nested within the broader urban backcloth, and how such facilities can influence risk at contrasting temporal scales. Utilizing Villa Park Stadium in Birmingham, England as a case, this paper seeks to identify the spatial distribution of pick-pocketing theft risk within a 3 km area around the stadium across various temporal periods on both sets of days. Police-recorded crime data and professional football match information for 104 games occurred during the 2005-2009 seasons were used for analysis. Results from the study demonstrate that crime in the area around Villa Park are significantly greater on match days than on controlled days, and that the crime events are not evenly distributed across whole time period, but more concentrated during the two hours prior to and one hour after the kick-off time. The regression findings show that the risk of pickpocketing theft is higher in pubs, bars etc, on match days than on control days, while the risk of pickpocketing theft on the control day is mainly concentrated in entertainment and other conventional crime hot spots. On match days, more pick-pocketing theft occurred on the places where more football funs gathering before kick-off, but after kick-off the crime risk shift to the pubs, bars, etc, which showing the crime pattern changed as time elapse from before kick-off to after kick-off. The research provides further support to ecological theories of crime and security and offers practical guidance for police and security in relation to where, when, and how to best allocate resources in the context of stadium-related mass gatherings.

Key words: football match, pick-pocketing theft, spatial distribution, negative binomial regression model, permutation test

摘要：

现阶段，关于足球比赛等体育活动影响周边区域犯罪风险的研究主要关注于一般性验证，而缺少对比赛期间不同时间范围内犯罪风险的态势分析及相关空间影响因素的研究。本文以英国伯明翰市的维拉公园球场为例开展实证研究，对2005—2009年赛季的104场职业足球比赛在比赛日和对照日、开赛前和开赛后等不同时间范围内球场周边3 km区域内发生的扒窃类案件的空间分布特征进行了分析，并利用负二项回归模型研究了相应案件的空间影响要素。结果表明：维拉公园球场周边区域内比赛日期间的扒窃案件发案量要明显高于对照日，而在具体的时间分布上，比赛日开赛时间前2小时到开赛后1小时内发生的扒窃案件显著偏多，并且案件在比赛日和对照日、比赛日的开赛前和开赛后期间的空间分布特征明显不同；回归分析结果显示，比赛日期间餐饮消费类场所的扒窃案件风险较为突出，且高于对照日，而对照日期间的扒窃案件风险则主要集中于休闲娱乐等常规性犯罪热点场所；在比赛日的开赛前，扒窃案件风险主要分布于一些人群聚集的公共场所，而在开赛后则主要集中于餐饮类场所，且随着比赛的开始扒窃案件的风险有转移的趋势。本研究对开展基于大数据分析的大型体育活动的犯罪风险精准防控具有一定的指导意义。

关键词: 足球比赛, 扒窃案件, 空间分布, 负二项回归模型, 置换检验