影響政府公共支出決策因素之研究

On Fuzzy Approach to Multi-AHP for Public Expenditure

吳文弘(Wen-Hung Wu) ； 盧惠伶(Fei-Ling Lu) ； 李仁棻(Ren-Fen Lee)

《商業現代化學刊》 5卷1期 (2009/03) Pp. 43-54

https://doi.org/10.6132/JCM.2009.5.1.04

摘要

本研究旨在衡量政府公共支出決策，受非理性之政治及其他因素影響程度。利用立意抽樣方法找到合適的專家，以尋求專家的反應，作爲合理之公共資源配置策略之參考。過程中以層級程序分析法(AHP)時，由於前後兩層級間加入中介變項，專家之反應可能會產生偏差，因此其權數之計算運用模糊中綜合評判方法，採用模糊合成法，以緩和級增加後產生之干擾狀況。並將產生之結果建立模型作爲預測之準則。透過殘差的檢視，省略與大多數專家共識差異很大的觀測值，使所建立的預測模型更加合適。本研究以我國中央政府爲例，建構迴歸分析模型，預測政府在財源不同因素影響下，合理公共支出之決策，並將通過一致性考驗之專家意見，作爲結論與建議之參考。

關鍵字

公共支出 ； 層級程序分析法 ； 迴歸分析 ； 立意抽樣

並列摘要

In view of continuing expansion in fiscal deficits in Taiwan government, this study explores the determinants affecting the expansion of public expenditures. First, we employ the Analytical Hierarchy Process (AHP) to set up a reasonable allocation strategy for public resources. While experts in different levels may make erroneous judgment, we use the comprehensive judgment technique in determining the weights and employ the fuzzy synthesis method to solve the hierarchical fuzzy issue. After analyzing the public expenditures of Taiwan's central government, we find that the expansion in public expenditures, along with its budget and implementation, is affected by the political changes, extrogenous environment changes, catastrophes, and financial sources. And the most serious deviation from the reasonable expenditure is the expenditures for economic development, next by the expenditures for science, education and culture, and then the expenditures for national defense. The least deviation in expenditures is the expenditures for social welfare.

並列關鍵字

Public Expenditure ； Analytical Hierarchy Process AHP ； Regression model ； Judgement Sampling