An Analysis on the Driving Factors of Population Changes in Local Provinces: Clues from News Big Data Analytics and Evidence from Empirical Analysis

Abstract

This article aims to reveal the driving factors of population changes in local provinces in South Korea based on both news big data analytics and empirical analysis. Even though many studies have focused on unveiling reasons behind population declines in local provinces, there is still a lack of empirical analysis investigating the explanatory powers of the potential factors of population declines or changes. To contribute to the literature on population declines and to the sustainable development of South Korea, we conduct amalgamated research composed of LDA on news articles and empirical analysis based on pooled regression models. First, by analyzing 8,699 news articles and 175,793 comments covering population declines, we unveil that there are five important topics about population declines and comments: Health, safety, finance & job, education, and culture, environment, & community values. Then, by estimating pooled regression models, we revealed that financial independent rate, EQ-5D (EuroQol-5D), the number of kindergartens, preschools, and continuing education facilities, environmental satisfaction, and community values are statistically associated with population change rates. At the same time, heterogeneous impacts of independent variables including the number of senior leisure facilities and kindergartens, and the influences of community values are estimated differently by 17 provinces.