交叉壓力、意見表達與政黨認同：2008年立法委員選舉的實證分析

Abstract

本研究旨在探討，在不同的人際互動關係之下，是否會影響民眾的政治表達？不同於以往研究政治參與偏重的「社會人口因素」（sociodemographic factor）與「政治動員」（political mobilization）等解釋途徑，本文援引「交叉壓力」（cross-pressures）作為研究架構，分析在2008年立法委員競選期間，個人處於交叉壓力之下，特別是自己的政治立場與其他家人不同時，對於意見表達的影響。作者採用「2005年至2008年『選舉與民主化調查』四年期研究規劃（III）：2008年立法委員選舉面訪案」（TEDS2008L）資料，將民眾區分為四種類型，以「政黨認同」（party identification）作為「意見表達」的代理變數，利用「雙變數交叉分析」（cross-tabulation analysis）與「多項勝算對數模型」（multinomial logit model）進行檢證，探討四種類型選民與政黨認同意見表達的相關性。研究結果顯示，在考量其他變數的效應之下，四種類型民眾與意見表達的關係，大致相符研究預期。換言之，儘管若干民眾抱持特定政治偏好與立場，然而受到交叉壓力的影響，因此在面訪調查時，往往委婉拒絕表達政治態度或者表示其政黨偏好，並回答不支持任何朝野政黨，而被歸類為屬於政治中立的獨立選民。這印證本文所提出的研究假設：在社會人際網絡中，較能處理交叉壓力的選民，其從事政治表達的機會較高；反之，較無法處理交叉壓力的選民，其直接表達政治立場的可能性較低。

This study examines whether people`s interpersonal relationship and social environments are important resources, which affect individuals` opinion expressions. Different from the approaches such of ”socio-demographic factors” and ”political mobilization” in the existing literature on political participation, this study uses ”cross-pressures” as its theoretical framework, analyzing the influence posed by cross-pressures on persons` inclination to express their own political opinions in the 2008 Legislative Yuan elections. The cross-pressures might be especially salient when their political positions are different from those of their family members. We employ the 2008 Taiwan Election and Democratization Study (TEDS 2008) survey data, using ”cross-tabulation analyses” and ”multinomial logit models” to investigate the association between four-type respondents and political expressions. Note that the variable of ”party identification” is employed as the proxy for tapping the concept of political expression. The findings reveal that the variables of gender, age, ethnicity, Taiwanese/Chinese consciousness, unification/independence preference, cross-pressures, and four-type respondents are the significant factors associated with an individual’s expression of party identification, and most of them run in the expected directions. More importantly, some respondents refuse to express their real partisan preferences and therefore report themselves as ”independents” during face-to-face interviews when they are under social interactions with cross-pressures. The results confirm the major hypothesis of this research: individuals having homogeneous social interactions and under low-level cross-pressures are predisposed to talk about politics, while people in cross-pressures involving greater political disagreement are less likely to express their political preferences.