Exploring the Use of Performance Information in Municipal Governments —Modeling from the Two-Stage Heckman Selection Estimation

地方政府運用績效評估資訊之探討－－使用二階段 Heckman 篩選模型進行分析

績效資訊係由系統化的測量設計所建構，用以提升公共服務的產能、效率及回應性。在實務上，績效評估資訊常與預算、策略規劃以及政府資訊作連結，用以檢視公共服務的成本、調整長程的規劃，並促進政府與民眾的溝通。地方政府運用績效評估資訊之動機究竟為何？ 本研究探討影響績效資訊運用的相關變項，並試圖解釋績效資訊的影響因子。分析對象採用美國「Stateof the Profession」(2006) 調查資料庫，有效問卷樣本數為2870，並運用二階段Heckman probit 篩選模型以避免採樣誤差 (sampleselection bias)。第一階段統計分析納入績效評估的背景變項，用以預測地方政府的政策採用行為；第二階段統計分析運用改良式的probit 模型，用以檢驗各假設變項與績效資訊用途的關係。分析結果指出，「策略計劃的實際操作」以及「對公民負責」這兩個變項對績效資訊的運用有正面的影響；然而，基於個人利己主義及有限理性的關係，「策略規劃的參與者」對績效資訊的運用反而有負向的影響。Performance information is constructed by systematic-measuring designs toward productivity, effectiveness and responsiveness. In practice, performance information could be linked with the budgeting process, strategic planning and citizen information; whereas decision-makers can undertake cost reduction, long-term planning as well as citizen-government communication. Why do municipal governments use performance information? This paper attempts to answer this question by examining factors affecting the application of performance information in the U.S. municipal governments. Data are from International City/County Management Association (ICMA) “State of the Profession” survey in 2006. The two-stage Heckman probit selection models are employed to avoid sample selection bias. The first step is a selection procedure which takes into account the adoption factors to estimate performance measurement adoption. The second step uses a corrected probit model to estimate the factor effect on performance information. The statistical results show that operation of strategic planning and concerns for citizen accountability have a positive and significant effect on the application of performance information. Due to individual egoist and bounded rationality, participants in strategic planning negatively affect the application of performance information.