選區、政黨與立法委員的三角關係---選制變遷前後的比較

Abstract

誠如Mayhew 所言，以當選連任為主要目的的國會議員，其在國會裡的行為必定與選舉的因素有關，而影響選舉的兩個最重要因素--政黨與選區，就成為國會議員在執行其代表職務時所面臨的兩股力量，而此兩股力量的相對大小，就相當程度決定著國會議員的代表角色與行為。長久以來立法委員選舉是以複數選區單記不可讓渡投票制（SNTV）為主，在SNTV 制下，立委有強烈的動機去建立個人選票，亦即讓選民投票的時候是基於立委本身的特質而非政黨的原因。為了達到此一目標，立委會用各式各樣的方式來表現自我、宣稱自己的功勞，諸如：在選區裡積極地服務選區民眾，在立法過程中積極的參與，利用機會帶特殊的好處到選區。甚至於，當選區利益與政黨的立場不一致時，立委即便不一定公然反對政黨，但是極可能消極的不支持（如不出席或不投票）。在2005 年6 月7 日通過的修憲案將立委選舉制度改為單一選區相對多數與政黨比例代表的混合制。過去對於立法委員角色行為的研究，大多以立委在SNTV 制下的選舉考量作為分析的基礎。那麼，當選舉制度變革之後，立委的角色行為是否也會跟著轉變？又如何轉變？究竟立委與選區、政黨的三角關係在選制變革之後會有如何的變化？在選制變遷之後，立委在選舉中不須像在過去一般面對同黨候選人的競爭，那麼，為了獲得選舉的勝算，是否會有與過去不同的爭取政黨提名與爭取選票的策略？而此一策略反映在他的代表角色與行為上，是否會與過去有所不同？此在在須要作有系統的研究方能解答。本研究將以第五至第七屆立法院為研究焦點，具體的研究時點係從2002 年到2009 年，如此方能充分掌握立委在舊選舉制度（第五屆）到新選舉制度（第七屆）之下的轉變。

When Legislators make decisions in the political process, they may face political pressures from two major forces. One force is their party and the other is their constituency. These two forces have profound influence on legislators』representative styles and behavior. Under a single non-transferable vote system, Taiwanese legislators have strong incentives to build the base of personal vote. Legislators thus try to attract electoral support through their personal qualities, qualifications, activities, and records, rather than through their party』s reputation. When legislators pursue such a strategy, they will tend to be more considerate of their constituency, conducting casework or supporting the adoption of pork barrel projects in exchange for the electoral support of their constituents. Whenever a conflict between legislators』party and their constituency occurs, they would tend to display more loyalty towards their constituency, rather than to their party. Even though they seldom oppose their party in public, they do not follow the party line. In June 2005, the electoral system had reformed from SNTV to a mixed system of single member district plural system plus a closed-listed PR system. Then, do legislators』 representative styles and behavior change because of the reform of electoral system? If they do change, then how do they change? This project will focus on a longitudinal analysis from the fifth term (when SNTV applied) to the seventh term (when the mixed system applied) to answer the previous questions. The research period of time ranges from February 2002 to July 2009.