An Analysis of Regional Differences in Consumption Expenditure Levels in China Based on Hierarchical Clustering

Abstract

There are significant regional disparities in the consumption expenditure levels of urban residents in China. Scientifically identifying the regional differentiation characteristics of consumption structure is of great significance for formulating differentiated regional development policies. Based on the data of per capita consumption expenditure of urban residents across eight major categories in 31 provinces, autonomous regions, and municipalities from the China Statistical Yearbook 2024, this study conducts an empirical analysis of regional differences in consumption expenditure levels in China using the systematic clustering method, specifically the within-cluster sum of squares method and squared Euclidean distance. First, descriptive statistics and boxplots are employed to identify the distribution characteristics of variables and detect outliers. Second, Pearson correlation coefficients are used to examine the relationships among variables. Subsequently, systematic clustering based on the within-cluster sum of squares method is performed, and the optimal number of clusters is determined with the aid of scree plots and dendrograms. Finally, one-way analysis of variance (ANOVA) is applied to test the statistical significance of the clustering results. The empirical results indicate that the 31 provinces in China can be classified into three consumption tiers. The first category is characterized as “high consumption–comprehensive development,” including Beijing, Shanghai, and Zhejiang, where the consumption structure has shifted toward development- and enjoyment-oriented consumption. The second category is defined as “moderate consumption–structural optimization,” covering Tianjin, Inner Mongolia, Jiangsu, Fujian, Guangdong, and Tibet, where the consumption structure is in a transitional stage of upgrading. The third category is described as “relatively low consumption–survival-oriented,” comprising the remaining 22 provinces, where consumption remains concentrated on essential goods. Significant inter-group differences are observed across all eight categories of consumption expenditure among the three groups (p < 0.01), and the clustering results are highly consistent with China’s regional economic development patterns. This study provides empirical evidence for understanding regional differences in consumption structure in China and for formulating differentiated consumption promotion policies.