문화소비지출에 미치는 변수들의 효과에 관한 연구: 사회, 문화, 경제, 및 정보자본을 중심으로

This study investigates the effect of variables within four-categories of cultural capitals on cultural consumption, including other leisure activities. In current literature, researchers analyzed cultural consumption using Bourdie(1984)'s homology and Peter and Kern (1992)'s Omnivore theories. Even though these theories contribute to current literature, they still lack many elements to explain current cultural consumption. Bourdie believes that the degree to which social origin and class distinction affects the preference when making ordinary choices. Peter and Kern mention that omnivorous consumers have an increased breadth of cultural consumption. We try to develop not only these two theoretical concepts, but also we use four categories of variables: social, economic, cultural, and information capitals. We collected 541 samples from online and offline sources after we controlled them by age and gender groups. Even though many current studies use the ordinary least square (OLS) method, we examined the collected samples with the quantile regression method since the variables in the four categories are heterogeneous. OLS can only show the conditional mean value and this tool is suitable for the central tendency distribution, but this method cannot explain the variations of cultural consumption. Thus, the OLS method is not enough for measuring cultural consumption. However, this study employs the quantile regression method, suitable in analyzing the heterogeneity in cultural consumption. The quantile regression method provides us with the values of cultural consumption at the different quantile levels. We noted that the four categories affect cultural consumption. In the case of social capital, two variables, age and status, were positive to cultural consumption using OLS. However, age, status, number of family members, and current living area were shown to be positive to cultural consumption using quantile regression. More specifically, current living area affects cultural consumption from to quantile levels. Also, the number of family members is positive to cultural consumption above quantile levels. In the case of cultural capital, the level of mother's and one's own education has a statistically positive impact on the cultural consumption at almost all quantile regression levels. For the economic capital, the income variable affects the cultural consumption at all quantile levels. The impact of the income variable in the middle to upper quantile levels is stronger than in the lower quantile levels. As for information capital, even though the online variable is statistically significant with the OLS method, the parameter shown to have positive impact only on cultural consumption from to levels using quantile regression. This study contributes to the current literature. Previously, cultural consumption was limited to the field of arts, but we decided to broaden the definition. To do so, we separated arts into high-level art and pop art, and added 3 more categories: tourism, sports, and activity clubs. Furthermore, the advanced statistical model, the quantile regression method, provides us with more specific results compared to OLS. In the future, we will try to investigate the relationship between the four capitals and specific cultural consumption. The results can provide us with the specific effects of the each variable on the four capitals such as social capital, economic capital, cultural capital, and information capital.