警察巡邏勤務科技創新之政策趨同與擴散：結合社會網絡分析與質化途徑

Policy Convergence and Diffusion of Police Patrol Technology Innovation: A Combination of Social Network Analysis and Qualitative Approaches

施沁懷(Chin-Huai Shih)

指導教授 ： 黃心怡

國立臺灣大學/社會科學院/公共事務研究所/碩士(2022年)

https://doi.org/10.6342/NTU202203742

摘要

本文以政策擴散之相關理論作為理論根基，探究於2019年年底至2021年5月期間，在警政署沒有給予任何誘因、要求與協助之情況下，仍有部分地方縣市警察局決定自主建置電子巡邏簽章（電子巡簽）系統的現象成因，並且特別關注縣市間的採用決策，是否與為何存在彼此相互連動與影響之關係。
在研究設計上，本文嘗試結合社會網絡之學說與分析技術，運用地理距離、跨縣市案件合作與警官人事調動等次級資料，建構與學習、仿效與競爭之政策擴散機制具有緊密理論關聯性的網絡連帶與結構變數，並使用對偶分析（Dyadic Analysis）驗證這些變數對於各縣市電子巡簽採用決策的影響。此外，本文更進一步藉由訪談資料強化各項網絡變數與特定擴散機制之連結，並就量化分析無法解釋的問題提供進一步的回應。
主要發現如下：(1) 網絡變數大幅提升分析模型的預測準確性，顯示存在明顯的擴散現象；(2) 各縣市在創新上的競爭意識與相對剝奪感，是電子巡簽系統最重要的擴散動力來源，其主要體現在先後採用系統之縣市在協力網絡中的高結構對等（structural equivalence）關係，以及異黨籍縣市首長間的政治競爭壓力；(3) 地理鄰近性仍是影響跨縣市觀摩學習便利性的重要擴散因素，合作頻率與人事往來等社會鄰近指標則不然；(4) 網絡中心領先縣市的政策採用行為，未引致所有邊陲縣市的仿效，前者的示範效果受到地理鄰近性所調節，蓋決策者傾向認定鄰近縣市的政策具備更高的複製可能性。
本文的研究價值在於跳脫我國政策擴散文獻對於「政策學習」的長期關注，闡述如何藉助社會網絡分析以區辨學習、仿效與競爭之不同擴散機制在擴散過程中的作用與相對重要性。上述成果亦可作為警政署未來推動地方警政創新的參考。

關鍵字

巡邏勤務 ； 政策創新 ； 政策擴散 ； 對偶分析 ； 社會網絡分析 ； 混合研究法

並列摘要

This thesis studies the policy diffusion of the adoption of the electronic patrol signature systems by several local cities in Taiwan between late 2019 and May 2021. As those local police departments voluntarily adopted the new digital tools without receiving any requests, incentives, or assistance from the National Police Agency, it is worthwhile focusing on the interdependent relations of the adoption decisions among local city departments.
As for the research design, this research tests policy diffusion theory and social network theories by collecting secondary data to construct a dataset of the policy diffusion for the electronic patrol signature systems. Measures include geographical distance, cross-jurisdiction case cooperation, and personnel transfer records to construct relational and structural network variables that have close theoretical linkages to the policy diffusion mechanisms of learning, imitation, and competition. The Dyadic Analysis is performed to test these variables' influence on each city's adoption decisions. Additionally, interview data were collected to reinforce the understanding of the relation between network measures and different diffusion mechanisms derived from the quantitative findings.
Major findings are listed below: (1) Network measures greatly improve the predictability of the analytical models, suggesting a strong link between network relations and policy diffusion. (2) The sense of competition and relative deprivation between cities are the main driving forces in the diffusion process of the system. City departments with high structural equivalence in the collaborative network and with political competition between mayors from different parties are more likely to follow adoptions. (3) Geographical proximity is still an important factor that affects the convenience of observation and learning across cities, while social proximity indicators, such as collaboration frequency and personnel exchanges, are not significant factors. (4) The adoption decisions of leading cities in the center of the network don't have a direct effect on all the marginal cities. The former's demonstration effect is moderated by geographical proximity because decision-makers tend to assume the policy of a nearby jurisdiction has a higher duplicability in their own jurisdiction.
The contribution of this article is that it moves from the traditional accentuation of the local policy diffusion research community on "policy learning", but illustrates how to separate or discern the functions and relative importance of learning, imitation, and competition mechanisms in the diffusion process with the aid of social network analysis. The research could also serve as a reference for the National Police Department to promote policing innovations among local police departments in the future.

並列關鍵字

patrol duty ； policy innovation ； policy diffusion ； dyadic analysis ； social network analysis ； mixed-methods research

參考文獻

王立恒，2017，〈【e政府創新應用】警察執勤全面行動化，手機成敏捷辦案必需品〉，iThome，取自：https://www.ithome.com.tw/article/117303，檢閱日期：2021/7/31。

王光旭，2015，〈社會網絡分析在公共行政領域研究的應用〉，《調查研究—方法與應用》，34: 67-134。

王捷，2020，〈從超商到賴神的家 台南警巡邏簽到仍用紙本〉，自由時報，取自：https://news.ltn.com.tw/news/society/breakingnews/3090342，檢閱日期：2021/7/29。

朱宣怡，2019，〈2019智慧城市展─智慧警政是未來城市安全的關鍵〉，《警光》，755: 54-63。

呂妍庭，2020，〈巡邏箱智慧化 也有城鄉差距〉，中時電子報，取自：https://reurl.cc/RbM4Q9，檢閱日期，2021/12/4。