Correlation and regression analysis of the population change index of border municipalities of the Republic of Serbia, 2011–2022

Abstract

The border municipalities of the Republic of Serbia are most often seen as peripheral in the geographical, demographic, social and economic sense. Serbia, like other post-Yugoslav countries, underwent crucial changes during the 1990s. They encouraged the intensification of unequal regional development in the centre–periphery relation. Although some border territories of European countries deviate from this pattern, the border effect has led to numerous negative consequences of the polarization process in Serbia, especially in those border municipalities with a longer peripheral status. Since depopulation is one of the biggest development problems in Serbia, the population change index 2011–2022 was studied using methods of correlation and regression analysis of demographic and social-economic indicators. Correlation analysis found that there is a medium or strong relationship with six indicators, while regression analysis showed that even those negligible variables in correlation analysis, such as the number of employees in the primary sector, can be significant. By applying the multiple regression method, it was determined that the population change index of 44 border municipalities can be predicted using three independent variables: the rate of natural increase, the number of marriages and the number of employees in the primary sector. The obtained regression model proved to be significant, considering the satisfactory value of the adjusted coefficient of multiple determination (adj R²=0.707) and the p-values of the independent variables below the threshold value (p < 0.05). As an outlier, the municipality of Čajetina stands out, for which the adopted model predicted a decrease in the number of inhabitants by 12.8% (the actual decrease is about 1.1%), while the absolute residual values for the other border municipalities range in the interval of 0.05 (City of Sombor) to 7.83 (municipality of Tutin). The aim of the paper is reflected in the importance of the application of selected statistical methods in demographic analysis.