정당 선호와 정치적 세련도: 2016년 제20대 국회의원선거를 중심으로

Abstract

본 연구는 유권자의 태도(attitude)와 행태(behavior)에 대한 전통적인 정당일체감의 강고한 영향력을 전제하면서, 이와 별개로 정당호감도(feeling thermometer) 척도를 통해 측정하고 조작할 수 있는 단기적⋅상대적 정당 선호가 과연 독립적인 영향을 가지고 있는지, 그리고 한 걸음 더 나아가 그 영향이 정치적 세련도(political sophistication)에 의해 어떻게 조절되는지를 2016년 제20대 국회의원선거를 사례로 검증하였다. 분석 결과, 정치적 이슈 - 공천 과정 평가 및 대통령 국정운영 평가 - 에 대한 태도의 경우 정치적 세련도가 해당 이슈에 대한 당파적 선호의 영향을 조절하기 위해서는 이슈의 성격이 또 다른 중요한 조건이 된다는 점이 새롭게 밝혀졌다. 한편, 투표 행태에 대한 분석에서는 정치적 세련도의 조절 효과(moderating effect)가 다양한 상황적⋅제도적 조건들의 결합에 따라 당파적 선호의 영향을 강화하는 쪽으로 발현될 수도 있고, 정반대로 그 영향을 완화하는 쪽으로 발현될 수도 있다는 것이 확인되었다. 본 연구의 함의는 다음과 같다. 첫째, 본 연구는 정당일체감과 정당호감도는 공통적으로 당파적 선호의 범주에 속하지만 질적으로 상이한 개념이며 양자 모두 독립적인 영향을 가지고 있다는 기존의 이론적⋅경험적 논의를 재확인했고, 그 과정에서 순수 무당파(pure independent) 또한 단기적 내지 일시적으로는 특정한 당파적 선호를 가지고 있기 마련임을 새롭게 조명하였다. 둘째, 본 연구는 한국 정치학계에서 그동안 크게 주목받지 못해왔던 정치적 세련도의 조절 효과(moderating effect)를 인지적 차원(cognitive dimension)에 속하는 정치적 이슈 태도와 행태적 차원(behavior dimension)에 속하는 투표 선택을 대상으로 폭넓게 검증하였다. 셋째, 본 연구는 컨버스(Converse 1964)가 개념화 한 신념체계(belief system)의 이론적 한계를 보완하는 경험적 근거가 될 수 있다. 컨버스의 주장대로 정치적 세련도가 높은 유권자들의 내면에는 분명 체계적인 인지구조가 존재한다. 그러나 그 구조는 인간 이성에 토대를 둔 이념적 집적연관(ideological constraints)에 국한되는 것이 아니라, 각종 정치적 대상(political objects)에 대한 정서적 선호와 인지적 평가가 복합된 총체적 태도들로 구성되어 있는 것이다.

This study aims to understand the moderating effects of the political sophistication in partisan attitudes in the context of South Korea. Based on several theoretical and empirical analyses, which argue feeling thermometer ratings toward political parties differ from party identification in the "Michigan model" in many ways such as stability or variability, this study examines two points

first, whether short-term and relative party preference which could be measured by feeling thermometer has an independent influence on the political issue attitude and voting behavior, even when party identification's effect was controlled statistically, and second, whether its effects appear differently depending on the levels of political sophistication (i.e. an interaction between the party preference and the political sophistication). Empirically analyzing the survey data for 2016 National Assembly election in South Korea, this study examines political sophistication's moderating effects on the short-term and relative party preference which influences on voters' evaluations of the Saenuri Party's and the Democratic Party of Korea's candidate selection processes (a sort of 'easy issue'), their evaluation of President Park Geun-Hye's job performances (a kind of 'hard issue'), and their voting behavior. Analyses on political issue attitudes show that the political sophistication did not influence the evaluations of Saenuri Partys and the Democratic Party of Koreas candidate selection processes. However, the short-term and relative party preference exerted significant effect on the evaluation of latters processes, even when the party identification variable is controlled. Meanwhile voters' appraisal of Park Geun-Hye was influenced by the short-term and relative preference, interacting with the political sophistication level. To put it concretely, regarding the voters who preferred the Democratic Party of Korea during a short period of time, the higher sophistication level they had, the more negative their evaluations of Park Geun-Hye's performances were. In case of voting behavior, the short-term and relative party preference heavily influenced the vote choice even when the party identification's effect is controlled. In addition, regarding the voters in favor of the Democratic Party of Korea for a short time, the higher political sophistication level meant increased likelihood of voting for their preferred party and the candidate of their party

in contrast, the voters who preferred the People's Party for the moment rather showed more likelihood of voting for the Saenuri Party if they are more politically sophisticated. In other words, these results show that the influence of the short-term and relative party preference could be either strengthened or weakened by higher sophistication level depending on various conditions. This study has implications in several ways. First, with a novel way of variable operationalization and statistical analysis, it reaffirmed the existing theoretical and empirical discussion that the long-held emotional attachment (i.e. party identification) and short-term party preference are two qualitatively distinguishable partisan preferences. Second, this study explored the political sophistication's moderating effects on the short-term and relative partisan preference's influence in both cognitive dimension (i.e. political issue attitude) and behavioral dimension (i.e. vote choice), and went further to investigate its empirical evidence in the context of South Korea. Third, this study newly insisted that, regarding the political issue attitude, the characteristic of the very issue is one of the important conditions for the political sophistication to moderate the short-term and relative party preference's influence. Results imply that highly sophisticated voters could rationalize their party preference or are more likely to show selective information processing in favor of their preference. In another way, however, the more politically sophisticated voters are, the more consistent and systematic relationship between their party preference and attitude(or behavior) is. In other words, highly politically sophisticated voters rationally structuralize a great deal of political information and feed it back into their attitudes. Fourth, regarding the voting behavior, this study showed the possibility that the moderating effect of political sophistication could either strengthen of weaken the influence of the short-term and relative party preference, depending on the combination of multiple conditions. Although the conditions were not set forth from the research design with scientific criteria, this study would be a useful precedent to the following researches. Fifth, these results would provide empirical evidence which supplements the theorectical limitation of the 'belief system,' conceptualized by Philip Converse(1964). Like what he maintained, a systemic cognitive structure exists among the highly sophisticated voters. However, that structure is not confined to ideological systemicity based upon the human reason, but rather also consists of a general attitude, encompassing affective preference and cognitive appraisal toward the political objects.