Amidst complex global economic conditions and increasing downward pressure on the world economy, enhancing domestic trade has become crucial for many countries to ensure sustained economic development and reduce reliance on international trade. China, with its rapid economic growth and diverse regional development, has embraced a 'dual circulation' development paradigm prioritizing the domestic market. This highlights the significance of inter-provincial trade in driving China's economic growth. While existing research suggests a positive relationship between inter-provincial trade and economic growth, there's a lack of consensus on the mechanisms involved and regional heterogeneity. This paper aims to address these gaps by investigating the impact of inter-provincial trade on China's economic growth, exploring the underlying mechanisms (technology improvement and industrial structure upgrading), and analyzing regional heterogeneity in these mechanisms. The findings offer valuable insights for developing effective domestic trade policies.

Existing literature on the relationship between inter-provincial trade and economic growth in China is diverse in its approaches to data acquisition and analysis of the economic effects. Data acquisition methods include using official reports, model estimations (most widely used due to data limitations), and VAT invoice information. While most studies agree that inter-provincial trade promotes economic growth, there's inconsistency and a lack of research on the specific mechanisms. Some studies highlight the role of market integration and industrial structure upgrading, while others suggest a less definitive or even negative relationship. This study aims to clarify these inconsistencies by focusing on the mechanisms and regional heterogeneity.

This study utilizes panel data from 29 Chinese provinces (excluding Hainan, Tibet, Hong Kong, Macau, and Taiwan) from 2000 to 2021. The dependent variable is per capita GDP (or, in robustness checks, gross regional domestic product and per capita disposable income), representing economic growth. The independent variable, inter-provincial trade, is measured using a novel approach combining the transportation volume distribution coefficient method and the interregional gravity model to calculate inter-provincial trade flows. This approach considers supply, demand, and trade friction coefficients derived from railway freight data. The accuracy of the calculated inter-provincial trade data is validated by comparing it with total retail sales of consumer goods. Control variables include international trade (total import and export volume), technical input (domestic invention patent acceptance), investment (total fixed asset investment), and human capital (per capita years of education). A two-way fixed effects model is employed to account for unobserved provincial and time-specific effects. Robustness checks replace the dependent variable and independent variable with alternatives, and an endogeneity test is conducted using lagged inter-provincial trade as an instrumental variable. Furthermore, to examine the mechanism of action, mediating effects analysis using total factor productivity and the ratio of tertiary to secondary industry added value (as proxies for technology improvement and industrial structure upgrading, respectively) are applied separately for the entirety of provinces and separately for the East, Central, and West regions, respectively.

The baseline regression results show a significant positive relationship between inter-provincial trade and economic growth. A 1% increase in inter-provincial trade leads to a 0.19% increase in per capita GDP. Robustness tests using alternative measures of economic growth and inter-provincial trade confirm this positive relationship. Endogeneity tests, using lagged inter-provincial trade and additional instrumental variables related to domestic market convenience and constraints, also support the finding that the relationship is not driven by reverse causality. Further analysis reveals that the effect operates through two main mechanisms: technology improvement and industrial structure upgrading. However, these mechanisms show significant regional heterogeneity. Technology improvement is only statistically significant as a mediating mechanism in the central and western regions, while industrial structure upgrading is significant in the eastern and central regions but insignificant in the western region. The control variables (international trade, technical input, investment, and human capital) also show significant positive effects on economic growth, consistent with theoretical expectations.

The findings confirm the significant role of inter-provincial trade in promoting China's economic growth, supporting the policy focus on strengthening the domestic circulation system. The mechanism analysis reveals the importance of both technology improvement and industrial structure upgrading. However, the regional heterogeneity highlights the need for nuanced policy approaches. The eastern region, despite benefiting from inter-provincial trade, relies more on international trade for technological progress; thus, policies should focus on leveraging industrial structure upgrading. In contrast, central and western regions can benefit from policies that boost technology absorption and improvement via inter-provincial trade. The insignificant effect of industrial structure upgrading in the west may be due to underdeveloped infrastructure, unbalanced resource endowment, limited consumer demand, and industrial relocation issues. The results emphasize the importance of considering regional contexts when formulating trade policies.

This study provides strong empirical evidence for the positive impact of inter-provincial trade on China's economic growth, operating through technology improvement and industrial structure upgrading mechanisms with regional variations. The findings underscore the need for differentiated policy interventions tailored to the specific conditions of each region to maximize the benefits of inter-provincial trade. Future research could explore the impact of other forms of communication (digital trade) and examine regional heterogeneity in more detail, including comparisons between coastal and inland provinces or the impact of specific trade policies.

The study has some limitations related to data and methodology. The measurement of inter-provincial trade relies on estimates from the transportation volume distribution coefficient method and the interregional gravity model due to the absence of consistently defined official data. This estimation involves assumptions that may introduce error. Specifically, the estimation of the friction coefficient and the handling of net outflow data after 2017 may lead to some deviation from reality. Future research using more comprehensive and consistently defined data, such as VAT invoice data, would improve the accuracy of the results.