Multidimensional spatial inequality in China and its relationship with economic growth

Spatial inequality, a major form of social inequality, significantly hinders sustainable development and human well-being. While global efforts focus on reducing poverty, substantial disparities persist between and within nations. The UN's Sustainable Development Goal 10 aims to reduce these inequalities. Regional inequality manifests in various dimensions: economic development, access to education, healthcare, housing, infrastructure, and digital services, as well as environmental justice issues. The COVID-19 pandemic exacerbated these inequalities. This study aims to understand spatial inequality in China across multiple dimensions and its evolution with economic growth. Williamson's inverted U-shaped hypothesis, suggesting an initial widening and subsequent narrowing of regional economic disparities with economic development, has been extensively debated. While some studies support this hypothesis, others challenge its universality and applicability across various dimensions and contexts. China, despite rapid economic growth, continues to grapple with significant spatial inequality. Previous research has primarily focused on economic disparities, with limited multidimensional analyses. This study addresses these gaps by examining spatial inequality across five key dimensions in China, testing the applicability of Williamson's hypothesis beyond the economic domain, and utilizing spatial autocorrelation measures to analyze clustering patterns.

Existing literature primarily focuses on the economic aspects of spatial inequality, often examining income disparities using measures like the coefficient of variation, Gini coefficient, and Theil index. Studies have analyzed China's regional inequality, noting shifts from an East-West divide to a North-South disparity. Factors influencing regional economic disparities include geography, transportation, globalization, urbanization, technology, migration, and investment. The role of agglomeration economies, knowledge spillovers, and path dependency have also been explored, along with the impact of regional policies like the Western Development Strategy. However, research on non-economic dimensions like social, environmental, infrastructural, and innovation-related inequalities in China remains limited. The applicability of Williamson's inverted U-shaped hypothesis has been primarily tested in developed economies and within the economic domain. This study addresses this gap by examining the hypothesis across multiple dimensions in the context of China's rapid development.

This study uses data from 31 Chinese provincial units (excluding Hong Kong, Macau, and Taiwan) from 1990 to 2021. Thirteen indicators across five dimensions (economy, society, environment, infrastructure, and innovation) were selected, aligning with the SDGs and China's 14th Five-Year Plan. These indicators include GDP per capita, disposable income per capita, high school graduation rate, physician density, unemployment rate, PM2.5 concentration, carbon emissions per capita, granted patents per capita, R&D expenditure per capita, living space per capita, urban green space per capita, road density, and internet penetration rate. Data sources include various Chinese statistical yearbooks and other databases. To measure spatial inequality, the population-weighted coefficient of variation (WCV) and Gini coefficient were employed. Moran's I was used to assess spatial autocorrelation and clustering patterns. Regression models (linear, quadratic, and cubic) were used to investigate the relationship between spatial inequalities (using WCV, Gini coefficients, and Moran's I) and per capita GDP as a proxy for economic growth. The optimal model for each indicator was selected based on goodness-of-fit.

Analysis reveals that China exhibits the highest spatial inequality in innovation (patents and R&D expenditure) and the lowest in the social dimension. Since 1990, most indicators show a decreasing trend in spatial inequality, except for unemployment and carbon emissions, which increased. The economic and innovation gap between eastern coastal regions and central/western regions widened, while disparities in housing, healthcare, roads, and digitalization narrowed. The temporal evolution of the WCV for most indicators (GDP per capita, income, education, healthcare, roads, and internet penetration) follows an inverted U-shaped pattern, indicating that spatial inequality initially increases and then decreases with economic development. Unemployment rates exhibit a U-shaped pattern, while carbon emissions show an N-shaped pattern. Moran's I analysis reveals significant spatial clustering for PM2.5 and carbon emissions, indicating concentrated pollution. The spatial clustering of several indicators (GDP per capita, income, housing, carbon emissions, roads, and healthcare) also displays an inverted U-shaped trend. Growth rates vary across dimensions and regions, with eastern coastal provinces showing faster growth in economic and innovation indicators, while central and western regions exhibit higher growth rates in housing, healthcare, roads, and digitalization. The Northeast region lags behind in various development indicators, and its unemployment rate is increasing. Regression analysis using both WCV and Gini coefficient confirms the inverted U-shaped relationship for several indicators, with turning points generally below RMB 28,000 per capita GDP.

The findings support the extension of Williamson's inverted U-shaped hypothesis beyond the economic domain to several dimensions in the context of China. The observed decrease in spatial inequality for many indicators with economic growth suggests that continued economic development can contribute to reducing regional disparities. However, the rising inequality in unemployment and carbon emissions highlights the need for targeted interventions. The spatial concentration of innovation resources and infrastructure requires policy attention to prevent the widening of the digital divide and promote balanced development. The diverse growth rates across regions emphasize the need for regionally tailored policies.

This study provides a comprehensive analysis of multidimensional spatial inequality in China and its relationship with economic growth. The findings support the inverted U-shaped hypothesis for several indicators, but not for all, highlighting the complex interplay between economic development and spatial inequality across different dimensions. The study's limitations include data availability, particularly for some indicators with data starting only from 2000. Future research could explore the mechanisms underlying the observed patterns and investigate the generalizability of these findings to other countries and regions.

The study's reliance on macro-level data may obscure variations at more granular levels. The limited data availability for certain indicators before 2000 might influence the robustness of the inverted U-shaped hypothesis. The study focuses solely on China, limiting the generalizability of the findings to other contexts. Further research is needed to delve into the underlying mechanisms driving the observed patterns and the specific policy implications for different regions.