Page 129 - IJPS-11-5

International Journal of
Population Studies Regional disparities and fertility rates

Table 2 describes the variables used in the time series disparities, and fertility rates. Before conducting multiple
model constructed to analyze the impact of regional regression analyses, a correlation analysis was performed
disparities on fertility rates. The dependent variable was to examine the relationships among key variables. In this
the national total fertility rate for the corresponding year, study, Pearson’s correlation coefficients, which are widely
as provided by Statistics Korea. To assess the influence of used as statistical measures of correlation, were applied
economic growth on the fertility rate, the GRDP growth to assess the relationships between regional fertility rates,
rate for the respective year was utilized as a common regional GRDP, and regional disparities in individual
independent variable. The Gini coefficient and Theil variables.
coefficient served as indicators of inequality. To balance
the strengths and weaknesses of these two indices, the 2.2.2. Spatial panel model
Gini coefficient of GRDP was used in Models 2 and 4, The spatial dependence of fertility rates across various
while the Theil coefficient of GRDP was applied in Models regions has been highlighted in multiple studies (Brée &
3 and 5. By employing these coefficients, the net effect of Doignon, 2022; Campisi et al., 2020; Kim & Jun, 2021;
regional disparities on the fertility rate can be estimated, Vitali & Billari, 2018). This study investigated macro-level
thus answering research question 2. In addition, in Models factors that influence the total fertility rate across different
4 and 5, the previous year’s total fertility rate was included regions, not only through correlation analysis but also
as an independent variable to examine its relationship with from a spatial perspective. To this end, a panel dataset
fertility rates in subsequent years. combining cross-sectional and time-series data, along
with spatial panel models, was utilized with regions as the
2.2. Analytical models spatial units.

2.2.1. Bivariate model The panel model improves estimation efficiency
The primary objective of this study was to investigate by increasing degrees of freedom through sample
the relationship between economic growth, regional augmentation and mitigating multicollinearity among

Table 1. Variables used in the spatial panel model

Variables Description Model 1 Model 2
Dependent variable
Log (total fertility rate) Log (regional fertility rate) ◯ ◯
Independent variables
GRDP GRDP (unit: trillion won) ◯ ◯
Density Population density (unit: 1,000 people/km ) ◯
2
Population Total population (unit: million people) ◯
Youth Population of 25 – 34 years old (unit: 10,000 people) ◯
Unemployment Unemployment rate (unit: %) ◯
Source: Statistics Korea.
Abbreviation: GRDP: Gross regional domestic product.

Table 2. Variables used in the time series model

Variables Description Model 1 Model 2 Model 3 Model 4 Model 5
Dependent variable
Fertility rate National fertility rate ◯ ◯ ◯ ◯ ◯
Independent variables
Growth rate GDP growth rate ◯ ◯ ◯ ◯ ◯
Gini Gini coefficient of GRDP ◯ ◯
Theil Theil coefficient of GRDP ◯ ◯
Fertility rate_1 Fertility rate of the previous year ◯ ◯
Source: Statistics Korea, author calculations.
Abbreviations: GDP: Gross domestic product; GRDP: Gross regional domestic product.

Volume 11 Issue 5 (2025) 123 https://doi.org/10.36922/ijps.8157