Verification of Zone-30-policy effect on accident reduction using propensity score matching method for multiple treatments
Verification of Zone-30-policy effect on accident reduction using propensity score matching method for multiple treatments
复制标题
DOI：
10.1016/j.cstp.2021.03.007
复制
发表时间：
2021-05-22
影响因子：
2.5
通讯作者：
Inoue, Satoru
中科院分区：
文献类型：
--
作者：
Seya, Hajime;Yoshida, Kazuki;Inoue, Satoru
本研究分析了日本兵库县30区政策对减少交通事故的影响。具体而言，以兵库县微区(cho-cho-moku)为分析单元，采用多处理倾向得分匹配法，对2017年，即引入时速30公里后几年的横断面数据进行了统计验证。兵库县的微区分为三组：第一组没有30公里/小时的区域，第二组有30公里/小时的区域但没有物理装置，第三组有30公里/小时的区域和物理设备。初步分析结果如下：第3组的“死亡”或“重伤”(KSI)事故数量有低于第1组的趋势，而第1组和第2组之间没有显著差异。此外，第3组的交通事故总数不少于第1组。这些结果表明，在这一点上，将一个区域单独划定为30公里/小时区域对减少交通事故的作用可以忽略不计；相反，这些区域与物理装置相结合可以有效地减少KSI事故的数量。此外，作为一种验证方法，实证证明，利用两组以上的数据是衡量交通安全措施影响的一种有效方法。
In this study, the effect of the Zone 30 policy on traffic accident reduction in Hyogo prefecture, Japan, was analyzed. Specifically, with the micro-districts (cho-cho-moku) of Hyogo prefecture as the analysis units, statistical verification was performed for the cross-section data of 2017, i.e., a few years after the introduction of the "30-km/h zones," using the propensity score matching method for multiple treatments. The micro-districts of Hyogo prefecture were classified into three groups: Group 1 with no 30-km/h zones, Group 2 with 30-km/h zones but without physical devices, and Group 3 with 30-km/h zones and physical devices. The primary analysis results were as follows: The number of "Killed" or "Seriously Injured" (KSI) accidents in Group 3 tended to be lower than that in Group 1, whereas there were no significant differences between Groups 1 and 2. Moreover, the total number of traffic accidents in Group 3 was not less than that in Group 1. These results suggest that, at this point, designating an area as a 30-km/h zone, by itself, has negligible effect on reducing traffic accidents; conversely, such areas can effectively reduce the number of KSI accidents when combined with physical devices. In addition, as a verification method, it is empirically demonstrated that utilizing more than two groups is an effective approach for measuring the impact of traffic safety measures.