文章詳目資料

選舉研究 TSSCI

  • 加入收藏
  • 下載文章
篇名 超越藍綠? 政治版圖在2014 年臺北市長選舉的新應用
卷期 24:1
並列篇名 Transcending Ideological Barriers for Voting? New Applications of Political Territory Analysis in the 2014 Taipei City Mayoral Election
作者 張順全莊文忠
頁次 097-132
關鍵字 政黨認同政治極化Mann-Kendall 檢定法Theil-Sen 斜率 推估法變動點分析political party identitypolitical polarizationMann-Kendall testTheil-Sen slope estimatorchange-point detectionTSSCI
出刊日期 201705
DOI 10.6612/tjes.2017.24.01.97-132

中文摘要

2014 年臺北市長選舉由一位過去偏向綠營,但選戰中既標榜白色 力量,又不斷號召「One City One Family」的無黨籍候選人柯文哲醫師 最後以壓倒性票數勝出。因而,臺北市已打破傳統壁壘分明、根深蒂固 的藍綠政治版圖之說甚囂塵上。然而,「超越藍綠」出現在首善之都臺 北市究竟是迷思抑或是事實?本研究為深究其代表意義,乃以臺北市藍 綠政治版圖為基礎,引入多種分析趨勢變遷之無母數統計方法互為發 凡,提出多項統計證據佐證,並演繹其在臺北市長選舉的新應用。 申言之,本研究充分應用中央選舉委員會之總體選舉資料的特性, 以臺北市各里為分析單位,使用集群分析展現臺北市的政治版圖。其後 以Mann-Kendall 檢定法,據以定性分析國民黨候選人得票率在版圖內 的時空相對變動趨勢,並以Theil-Sen 斜率推估法進行趨勢斜率的定量 估算,再進一步評估變遷中顯著的改變點。最後以政治版圖為基礎,整 合取自總體選舉資料之衍生變數,得到代表特定政黨或藍綠陣營吸票 實力的相關訊息,同時觀察政黨催票的影響效果。研究結果除了發現 2014 年臺北市長選舉結果的確出現超越藍綠的現象,更進一步討論表 象背後真正的意涵。

英文摘要

The 2014 Taipei City mayoral election was overwhelmingly won by a grass-roots candidate, Dr. Wen-je Ko, who has been regarded as “Deep Green” in terms of Taiwan’s political spectrum, but promoted the slogan “One City One Family” for new white power in this campaign. The rooted voting behaviors described by traditional political territories in Taipei City were supposed to be on the verge of imminent collapse, and whether the signal of transcending ideological barriers for voting in this case was grounded in reality or myth, is still worth exploring. This study conducted several tests for trend analyses used in nonparametric statistics and interpreted new applications based on political territory in this mayoral election. Furthermore, this study has three primary innovative perspectives. First, based on the voting database from the Central Election Commission for election studies, this study can explain how to utilize the basic unit of household address, such as Li, for building political territories, and our models presented the grouping political spectrum structures of voting behaviors of voters living in Taipei City by clustering analysis. This study can also develop a more novel approach to the combination of domestic political territory research and trend analyses embedded in nonparametric statistics, such as the Mann-Kendall test, the Theil-Sen's slope estimator, and Pettit test statistics for change-point detection, which are all adapted to analyze the trend characteristics for changing voting behaviors in Taipei City. Finally, based on built political territories, and linked with relevant concepts of political polarization, political party identity, and allocation effects when political parties drum up votes, the research results can determine and gain insight into the transcending of ideological barriers for voting in the 2014 Taipei City mayoral election.

相關文獻