摘要 犯罪的管理，是政府治理上一項重要的課題。一個城市犯罪率的呈現，往往代表城市治理的重要指標。影響犯罪行為發生的因素很多，不同地區可能產生的外在環境因素會有不同，而犯罪手法、類型及防治的手段亦隨之相異。因此，有必要對於不同影響犯罪的成因予以探討。本研究除承續學者過去研究成果之外，還會建立可能產生犯罪影響的構面，藉由犯罪學理論的發展及實證研究，討論犯罪發生率與社會環境變動所具有的主要特徵間的關聯性。本研究係運用20年間，20個縣市的縱斷面資料，建構迴歸模型，檢驗各縣市人口結構特性、城市經濟活動指標及環境污染指標等三個構面與犯罪率的關係。研究結果發現，經濟失業率、二氧化硫含量均與犯罪呈正向關係；且縣市之自有住宅持有率比例愈高，則犯罪率愈低。本研究之統計分析，係透過擷取臺灣20縣市社會環境三項構面資料所完成的縱橫迴歸分析模型（Panel Regression Model）結果，檢視人口結構及外在環境特徵與犯罪率之關係，從而擬具治安策略建議，以提供政府相關機關作為擬訂犯罪預防政策及訂定打擊犯罪之治安策略作為之參考。;Abstract The administration of crime holds significant importance in government governance, with a city′s crime rate serving as a crucial indicator of urban governance. Numerous factors influence the occurrence of criminal behavior, and these factors can vary across different regions due to diverse external environmental influences. Consequently, the methods, types, and preventive measures for crime differ accordingly. There is a need to delve into the various causes of criminal activities. This study investigated factors that influence crime, utilizing the development of criminological theory and empirical research. This research employed longitudinal cross-sectional data spanning from 1998 to 2020, encompassing 20 counties and cities in Taiwan. A panel multi-regression model was constructed to assess the relationship between population structure characteristics, urban economic activity indicators, environmental pollution indicators, and the crime rate. The empirical results revealed a positive association between economic unemployment rates and sulfur dioxide content with crime. Additionally, a higher proportion of owner-occupied housing in a county or city was found to result in a lower crime rate. The findings of this study can aid in crafting recommendations for public security strategies, offering government agencies valuable insights for developing crime prevention policies and formulating effective public security strategies to combat criminal activities.
