SNTV選制下“立法委員連任勝選幅度模型”建立之嘗試－以第四屆區域立委為例

立法委員連任分析對國會制度化的研究，意義重大。但國內關於立法委員之連任分析研究，多仍僅停留在個案分析的階段，無法做一整體的模型建構，而原因在於SNTV選制中選區規模大小不一，使得跨選區的立委比較研究難以進行，尤其是在得票數的分析上。本研究嘗試運用族普基數，將得票數計算轉化成為勝選幅度，使不同選區立委之得票數得以立於客觀基準進行比較，並嘗試以此建立立法委員連任勝選幅度迴歸模型。經由檢驗後，發現模型相當穩定，且詮釋力也頗佳。而本模型亦顯示，政黨因素與個人表現因素兩者，同時為影響立法委員連任勝選幅度表現的關鍵因素。

關鍵字

SNTV選制；立法委員；連任；連任勝選幅度模型

並列摘要

The study of legislators' reelection is very significant for the research of the institutionalization of the parliament. But domestic research about legislators' reelections remained in case studies and am aggregative model could not be built due to the different scales of districts. This makes comparing legislators from different districts difficult, especially with regards to vote-getting. This research attempts to use Droop quota to translate vote-getting into ”reelection magnitude” in order to objectively compare the vote-getting of different legislators of different districts, and to build the regression model of their reelection. Through this examination, we find that the model is very stable and interpretive; meanwhile, the model reveals that both party and individual factors are critical in affecting fourth-term legislators' reelections in the fifth legislative election.