本研究旨在探討「環境效應」(contextual effect)對於選民投票行為的影響程度。迥異於以往國內學術文獻對於地理區域的分類方式(諸如:實際得票情形、行政區域、現代化程度,以及不同政府類型的地區等),本研究嘗試引介近年來美國政治學者探討種族議題所發展的「賦權理論」(empowerment theory),修正其概念以適用於台灣地區的特有環境,據此作為劃分地理區域的指標,並探討其對於2001年縣市長與立法委員選舉選民投票抉擇的影響。本研究採用「2001年台灣選舉與民主化調查研究:民國九十年立法委員選舉全國大型民意調查研究」資料,藉由「雙變數交叉分析」與「多項勝算對數模型」(multinomial logit model)進行檢證。研究結果顯示,在考量政黨認同與其他變數的影響之下,「政治賦權」(political empowerment)環境效應與選民投票行為之間仍然存在著顯著關係,這印證本文所提出的研究假設:民進黨賦權效應較高的區域,選民愈趨向支持民進黨候選人;反之,民進黨賦權效應較低的區域,選民則傾向支持國民黨與親民黨候選人。在結論中,本文摘述研究要點,比較「賦權理論」與「政治版圖」研究途徑的異同之處,並且提出未來研究的方向與建議。
This work aims at examining the contextual effects of ”political empowerment” upon voting behavior in Taiwan. Different from the regional classifications developed in preceding literature (e.g., vote percentage of party or candidate, administrative boundary, degree of modernization, or divided and unified government), this research employs and modifies ”empowerment theory” of the (ethnic/racial) minority politics to account for the shifting electoral fortunes of the Kuomingtang (KMT) and the Democratic Progressive Party (DPP). We take advantage of the 2001 Taiwan's Election and Democratization Study (TEDS 2001) survey data and examine whether different empowerment areas exert significant impacts on voting behavior in the 2001 elections to the Legislative Yuan and county magistrates and city mayors. The methodology adopted in this study involves two steps. The first approach is the use of cross-tabulation analyses, and the second method employs the multinomial logit model in order to evaluate the simultaneous effects of independent variables on the dependent variable. As hypothesized, the results indicate that the contextual effects of ”political empowerment” still emerge as statistically significant for accounting voting choices even as party identifications and other explaining variables are taken into account. The findings demonstrate that constituents in high-DDP-empowerment areas-as indicated by control of the mayor's/magistrate's office-tend to vote for DPP candidates than those living in low-DDP-empowerment areas, and vice versa. In the conclusion, we review the major findings and limitations of this study and compare the research approaches of ”empowerment theory” and ”political geography.”