犯罪群聚：2001~2008年台灣竊盜犯罪的空間分析

Abstract

本研究主要利用空間分析方法進行臺灣地區竊盜犯罪的分析，以日常活動理論為基礎，建立適當分析架構對利於犯罪行為之環境因素進行探討，並運用空間分析方法，以分局的空間分布資料進行比對，進一步確立分析架構的可行性，供未來犯罪防範之參考。 本研究的結論分成三項：第一，依空間分析探索可知，竊盜犯罪率呈現群聚於某些特定的地區，如臺中市、新竹市等都會地區，此外，臺中市第六分局轄區為西屯區的竊盜犯罪從2002年急遽地增加；然而臺南市第二分局是隨年度而逐漸往上攀升的現象臺中地區為眾所皆知的情色場所聚集處，又居臺灣中心位置，交通中樞點，潛在犯罪者實行犯罪行為後，具有較佳的銷贓管道，使得該地成為近年來竊盜犯罪率的高位。就Lisa圖而言，主要犯罪熱區有臺中市第四(五、六)分局、臺南市第三(六)分局、高雄市三民分局、新竹市第一分局。犯罪冷區為宜蘭縣三星分局、臺中縣和平分局、高雄縣六龜分局、臺東縣關山分局等非都會地區。 第二，根據傳統迴歸模型，社經變數達到顯著性的社經變數有「離婚率」、「人口密度」、「人均所得」及「地價指數」等變數，我們無法擔一個因素解釋一地竊盜犯罪率。 最後，控制各類社經因素後，以SLM迴歸模型知，在其他變數不變下，竊盜犯罪率具鄰近效應，並且受到空間鄰近變數影響，空間鄰近變數的係數值為0.4159，其影響方向為正，即以分局為單位的鄰居竊盜犯罪率會影響本地的竊盜犯罪情形。

In this paper, we had the routine activity theory as the basic ground to built up a suitable analysis framework, and adopted the Spatial Analysis as the main tool to make it a standard and to study the crime rate of theft from 2001 to 2008 in Taiwan. Then, we put the result of the Spatial Analysis and the information of areas of specific precinct together to get the difference between them. In this way, we would know whether this method works out or not, and become the useful data for improving crime rate in the future. There are three conclusions in this paper. First, based on analysis data of Exploratory Spatial, it showed that the crime rate of theft clustered in specific region, like Taichung City, Hsinchu city and so on. Besides, the crime rate of theft in Xitun District of Taichung Sixth Precinct had rapidly increased since 2002. In addition, after ranking years of the crime rate of theft in every precinct, the phenomona showed out that it increased rapidly in the Taichung Sixth Precinct while the Tainan Second Precinct went up gradually year by year. And why Taichung Sixth Precinct was on the top？It’s because Taiching has been famous for its pornographic place,and it’s located in the center of Taiwan.And because of it’s the transport hub, potential criminals would have much better pipeline to dispose stolen goods after commiting crimes.As for hotspots in the Lisa map,there are several cities with high crime rate like the Forth ( Fitfh, Sixth) Precinct of Taichung, the Third(Sixth) Precinct of Tainan, the First Precinct of Hsinchu. While the coldspots are Yilan Sansing Precinct, Taichung Heping Precinct, Kaohsiung Liou-Guei Precinct and Taitung Guanshan Precinct. The result conveyed that the low-low crime rate happened in those rural area. Second, according to the Empirical Regress Model, among the variables of economy and society, there were some reached significant effect like the divorce rate, popularity density , average income, the index of landprice and other variables. Above all, the behavior of crime rate could not be explained by only one factor. Finally, after putting each kind of social and economic factors into control, the crime rate of theft had neighborhood effect which meant it was influenced by spatial neighborhood variables. And its coefficient was 0.4159 in the SLM regress model. In other word,as for precint unit, we found that the crime rate of theft of neighborhood would affect native’s.