環境議題之比較社經風險分析及其健康風險共同因素探索－以台灣專家認知為例

Abstract

比較風險分析（comparative risk analysis）最先係由美國環保署所提出，用以評估環境問題潛在健康、生態風險及對社會經濟之衝擊，並應用在環境問題風險排序上，以作為解決環境問題優先度之參考。然而，傳統比較風險分析之排序方法，並未處理不同序位間之環境議題是否存在顯著差異的問題，故而影響資源分配之效益。為有效解決此一問題，本研究首先構建一比較風險分析研究模式，並應用此模式進行社會、經濟層面之比較風險分析實證研究。該模式由輸入、處理及輸出三階段組成，輸入階段包括環境議題與評估指標之界定以及風險認知資料收集；處理階段係利用重複測度分析與集群分析為環境議題之風險大小差異作檢定與分類；輸出階段則就分類結果進行風險大小排序。比較社會經濟風險實證結果顯示，環境議題之風險差異確因不同之評估指標而產生變化，排序結果已藉由圖表呈現。 由於環境議題造成之健康風險較其它二項衝擊更易被感受到且是最切身的問題，為更有效益地降低我國環境議題造成之健康危害，本研究第二部分由健康風險之相關性著手，試圖找出相關性高環境議題之共同因素，以便針對共同因素提出改善策略，這部分的實證研究是基於專家認知資料，以探索性因素分析獲取共同因素，並以文獻為依據進行致因探討。結果顯示，二十一項環境議題之共同因素，可歸納為環境基礎建設與管理不健全、空氣相關之議題、人類誘發之自然生態系統破壞、溫室氣體之排放、風險規避意識之缺乏及水資源管理不善。在風險管理上，我們建議針對此六項環境議題共同因素提出改善策略以獲取較佳之成效。

Comparative risk analysis is an evaluation process involving the relative ranking of specific environmental issues based on their estimated risk to human health, ecology and quality of life. In 1987, the U.S. Environmental Protection Agency (EPA) conducted the first comparative risk evaluation. However, the conventional approach of comparative risk analysis does not check if there exist significant differences among environmental issues and therefore resources cannot be effectively allocated. To deal with this shortcoming, we attempt to provide an effective statistical model for analyzing environmental issues and apply the model to an empirical study of comparative socioeconomic risk in Taiwan. The proposed statistical model consists of three stages: input, process and output. In the input stage, environmental issues whose perceived risks are to be compared need to be first identified. Evaluative criteria and other potential moderator variables also need to be determined. Then risk perception data are collected. In the process stage, two statistical methods, repeated measures analysis and cluster analysis, are used. Repeated measures analysis is used to test if risk difference will be affected by evaluative criteria and/or other moderator variables and if risk difference is significant. Cluster analysis is used to group environmental issues in such a way that environmental issues within a cluster have similar degree of perceived risks and different clusters of environmental issues have significantly different degree of perceived risks. Finally, in the output stage, risk-based priority setting is made for clusters obtained. The empirical results indicate that selected impact areas do affect relative risk differences among 24 environmental issues, and the difference is significant for each area. The rankings of 24 environmental issues are reported. Health risks are thought to have the most direct threat to human beings. To reduce health risks more effectively, we attempt to identify intercorrelated environmental risks based on experts’ perceptions and then explore their common factors. The common factors have been identified by using exploratory factor analysis with literature support. They are unsound environmental infrastructure and management, airborne issues, human-induced damage of natural ecosystem, greenhouse gases, lack of risk-avoidance sense, and lack of water resources management. In risk management, attention should be directed to these underlying dimensions so that environmental issues could be handled more effectively and economically. The approach presented and the findings of the study may provide some guidelines for risk management.