毒品使用及犯罪行為關聯性之研究-以P.E.S.模式分析為例

An Analysis of the Causation Relation between Drug Use and Crime - Illustrated by the Analysis of P.E.S. Model

顧以謙(Yi-chien Ku)

指導教授 ： 楊士隆

國立中正大學/教育學院/犯罪防治學系/博士(2016年)

摘要

過去研究指出毒品使用和犯罪行為存在三種因果關係模式：心理藥物模式（The psychopharmacological model）、經濟動機模式（Economic motivation model）、組織系統模式（The systemic model）。但台灣本土犯罪學尚無實證資料將此三大模型納入測試及分析，在解釋毒品及相關犯罪問題之實證研究上付之闕如。因此，本篇論文透過文獻探討整理過去有關毒品犯罪及相關犯罪問題，並以問卷實體調查監獄受刑人之毒品施用經驗與過去犯罪情形，建構模式探究毒品施用與犯罪問題之關聯性。本研究於2016年7月至8月前往台中監獄及台中女子監獄施測，共發出1,260份問卷，扣除遺漏、反應不佳等無效樣本後，共獲1,239份有效樣本，男女樣本回收率皆為98%左右，在95%信心水準下，抽樣誤差最大為正負2.72%。
本研究透過負二項迴歸模型(Negative Binomial Regression Model, NBR)分析發現，心理藥物模式方面，海洛因及安非他命成癮性對於施用海洛因後暴力行為具顯著正相關。大麻成癮性對於施用大麻後暴力行為雖具有顯著影響，然其相關性會受到憂鬱困擾之調節（moderated），顯示大麻成癮性對暴力行為發生之影響，在無憂鬱困擾者身上反應較有憂鬱困擾者強烈。後續分析中，本研究發現施用二級毒品對於財產犯罪具顯著影響，當施用二級毒品每增加一個單位（件），其發生財產犯罪行為之勝算比增加18.3%。再者，於經濟動機模式方面，冒險取向對海洛因成癮性影響暴力行為具正向調節作用。第三，海洛英、安非他命成癮性對於財產犯罪也具顯著預測力。第四，投機性（simple-task）對經濟壓力影響施用海洛因後財產犯罪行為之預測力具有正向調節作用。此外，投機性、冒險取向對經濟壓力影響施用大麻後財產犯罪行為預測力具正向調節作用。最後，在組織系統模式方面，研究結果指出曾因販賣海洛因、安非他命與其他藥頭起衝突佔整體約9~11%，引發流血事件者佔整體4.4%~4.9%。進一步分析，本研究發現偏差團體認同感對於財產犯罪行為具顯著正向影響，而偏差團體認同感更可強化社會孤立感對「暴力行為」之影響。當偏差團體認同感及社會孤立感之交互作用每增加一個單位（分），其發生暴力犯罪行為之勝算比會上升9.2%。
本研究建議在未來能進一步擴大調查規模及樣本數，將所有毒品種類及犯罪行為類型及發生時間順序劃分得更為細緻，對於未來建構毒品使用及犯罪行為模式將更有助益。此外，囿於時間資料取得不易，建議未來政府在妥適保護個人資料下，開放研究者申請個案犯案事件及時間等資料，以利未來研究發展。

關鍵字

毒品 ； 毒品關連性犯罪 ； 心理藥物模式 ； 經濟動機模式 ； 組織系統模式 ； 負二項式迴歸模型

並列摘要

An earlier research consolidated drug abuse and criminal behaviors into three cause-effect models: the psychopharmacological model, economic motivation model and systemic model. However, locally in Taiwan, these three models have not yet been incorporated into tests and analysis of empirical studies associated with criminology, nor interpretation of drugs or drug-related crimes. In view of above theory testing concern, this thesis aims to construct a model to explore the relationship between drug use and crime through review of research literature on drug related crimes and criminal issues, as well as an empirical survey-based study on the experiences and criminal offending profiles of a group of inmates incarcerated for drug related offending. A survey was implemented between the period from July to August of 2016 at Taichung Prison and Taichung Prison for Women. A total of 1,260 questionnaires were sent out and 1,239 were considered effective samples after eliminating questionnaires with omissions and poor responses. The return rates for both male and female groups fell in proximity of 98% and, with 95% of confidence level, the maximum sample error is +-2.72%. Negative Binomial Regression Model (NBR) was adopted to evaluate the connection between drug use and crime. Findings of this research are summarized by the three cause-effect models:
In the psychopharmacological model, addiction to heroin and amphetamine was found to have positive correlation with violence after use. Addiction to marijuana also has significant correlation to violence after use, but the degree of correlation is moderated by depression. Occurrence of violent behaviors associated with addiction to marijuana is less intense in persons without depression than persons with such mental conditions. Follow-up analysis also found that use of class 2 drugs is related to property crimes up to a significant level. When use of class 2 drugs rose by one point, the odds ratio of property crimes rose by 18.3%.
In the economic motivation model, the factor of “risk-seeking” is a positive moderator in the correlation between heroin addiction and violence. Also, addiction of heroin and amphetamine has predictive power to property crimes up to the significant level. “Simple-task” is a positive moderator to the predictive power of economic pressure in association with property crimes after use of heroin. Furthermore, both factors of “risk-seeking” and “simple-task” are positive moderators to the predictive power of economic pressure in association with property crimes after use of marijuana.
Finally, in the systemic model, findings of this research indicated that 9 to 11% of the respondents have been in conflict with drug dealers when selling heroin or amphetamine and 4.4 to 4.9% of the incidents ended with bloodshed. Further analysis indicated that identification to the delinquent groups is in a significantly positive correlation with property crimes. And such sense of identification reinforces the effect of social isolation on criminal behaviors. When the factor of “identification to delinquent group” rose by one point, the odds ratio of violent criminal behaviors rose by 9.2%.
This research recommends that a further research with expanded scale and sample can be implemented in the future. A large-scale study with more refined categories of drugs, criminal behaviors and time-sequence of occurrence will provide substantial aid to the establishing of a model for analysis of drug use and criminal behaviors. Furthermore, this research was limited by the difficulty of obtaining the time-related data. We recommend that, in the future, the administrative agencies may consider to open the data of criminal cases, including time-related information, under proper protection of personal information to researchers who seek in-depth study on individual cases. Certain degree of open data will facilitate further research and development.

並列關鍵字

drug ； drug-related crime ； the psychopharmacological model ； Economic motivation model ； the systemic model ； Negative Binomial Regression Model

參考文獻

李昆樺（2014）。大學生毒品使用之心理社會風險因子與模式驗證。國立成功大學，台南市。

楊士隆、曾淑萍、戴伸峰（2011）。台灣地區收容少年入院前非法藥物使用盛行率調查研究，青少年犯罪防治研究期刊，第3卷第2期，頁89-125

楊士隆、戴伸峰、曾淑萍（2011）。台灣成人受刑人入獄前非法藥物使用之盛行率調查，犯罪學期刊，第14卷第2期，頁99-137。

楊士隆、戴伸峰、曾淑萍、顧以謙、張梵盂（2015）。2015收容少年之非法藥物使用調查研究-以新北市、台中市、高雄市為例。第8卷第1期，頁162-22。

鍾鳳嬌、張立人、王國川（2015）。國中生參與廟會陣頭相關因素研究。人文社會科學研究，第9卷第1期，頁1-26。