政策擴散理論之探討–以反怠速政策為例

Using the Policy Diffusion Theory on the No-Idling Policy in Taiwan

吳偉豪(Wei-Hao Wu)

指導教授 ： 翁興利

共同指導教授 ： 李元和(Yuan-Ho Lee)

國立臺北大學/公共事務學院/公共行政暨政策學系/碩士(2013年)

摘要

臺灣地區近年來汽機車總量不斷攀升，相對二氧化碳等廢氣排放量也日漸增加。有
鑒於此，有關環境保護的政策也隨著民眾對於生活品質的需求提高而增加，反怠速政策
（No-Idling Policy）即為其中之一。然而，反怠速政策並非一開始就在臺灣地區全面實施，
而是由台南市率先實施以及部分縣市接連實施後，才由中央立法通過將實施範圍擴大至
全國。因此，本文探討影響反怠速政策擴散之因素為何。

本文是以「政策資訊擴散」為主要的參考架構，並結合空間迴歸模型之相關理論，
針對「反怠速政策」之採用者進行研究。而本研究目的如下：一、瞭解「反怠速政策」
的推行現況；二、利用政策擴散理論，探討相關文獻，找出影響反怠速政策擴散之關鍵
因素；三、透過空間迴歸模型分析，找出各縣市之間採用反怠速政策之影響因素；四、
針對分析結果，研擬改善建議，以提供未來進行反怠速政策相關研究之參考依據。

本研究發現如下：
一、 若以傳統迴歸模型分析，空氣中臭氧濃度、平均每人每年可支配所得對反怠速
政策影響具有正向顯著影響。
二、 在全域型空間自我相關Moran’s I 檢定，從2008 年至2011 年各縣市影響反怠速
政策之變數的Moran’s I 值大部分均大於0，代表各縣市相鄰地區具有空間的正向自
我相關。
三、 若以地域型空間自我相關分析LISA 檢定，由各縣市反怠速政策影響LISA 空間
分佈可發現，台北市與新北市對於鄰近縣市的反怠速政策影響具有正向關係影響，
而基隆市、台中市及嘉義市三個地區則有負向關係影響，意即台北市及新北市對於
反怠速政策實施會相互學習，基隆市、台中市及嘉義市則不會跟進實施政策。

關鍵字

反怠速政策；政策擴散；空間迴歸模型

並列摘要

In Taiwan, as the amount of motor vehicles increasing, the amount of carbon
dioxide and other exhaust emissions also increasing relatively in recent years. In view
of this, policies about environmental protection, as the public demand for improved
quality of life increased, and No-Idling Policy is one of them. However, No-Idling
Policy doesn’t implement comprehensively at the start in Taiwan, but by the first
implementation of Tainan City and some cities after the implementation of a series, it
will be implemented by the central legislation passed to extend to the whole country.
Therefore, this research examines the impact of No-Idling Policy diffusion factors.

This research is based on “Policy Information Diffusion” as the main reference
framework and combines with the theory of spatial regression model for “No-Idling
Policy” adopted by the study. The purposes of this research are as follows. First, to
understand the current situation of the implementation of “No-Idling Policy”; second,
use of policy diffusion theory and examine the relevant documents, to identify the key
factors of impact No-Idling Policy diffusion; third, using spatial regression analysis to
identify the impacting factors of other cities adopt No-Idling Policy; and last, to be
aimed at outcome, to investigate and plan forward for improving recommendations in
order to provide relevant No-Idling Policy studies, as a reference in the future.

The findings of this research are as follows. First, Based on the traditional
regression model, the ozone concentration in the air and the average disposable
income per person per year have positive influence on No-Idling Policy;second, with
Moran’s I test on the result of global spatial autocorrelation analysis, the Moran’s I
values of most variables of No-Idling Policy among counties are all showed greater
than zero in 2008 to 2011. These results indicated that it has a spatial positive
autocorrelation between cities or counties and their neighborhoods; and last, based on
the distribution of the Local Indicators of Spatial Association (LISA) of No-Idling Policy in all cities and counties, it found that Taipei City and New Taipei City
indicated a positive relations with their neighboring cities or counties, while Keelung,
Taichung and Chiayi City showed a negative relations with their neighboring cities or
counties, meaning Taipei City and New Taipei City for No-Idling Policy
implementation will learn from each other, but Keelung, Taichung and Chiayi City
will not follow up the implementation of policy.

並列關鍵字

No-Idling Policy；Policy Diffusion,；spatial regression model

參考文獻

李國榮，2010，《機車怠速熄火技術之研究》，屏東科技大學車輛工程系碩士論文。林淑

胡立諄、賴進貴，2006，〈臺灣女性癌症的空間分析〉，《台灣地理資訊學刊》，4：

許志義、王京明、郭婷瑋，2000，《臺灣二氧化碳減量政策之模擬分析》，中華經濟研究

黃東益，2004，〈全球治理下政府知識管理的新面向：府際政策學習〉，《國家政策季刊》，

楊碧蓮，2011，《影響地方政府環境績效因素之探討》，淡江大學會計學系碩士在職專班