TL;DR
This paper investigates how e-commerce, specifically online business and online shopping, spreads across China's counties, revealing distinct spatial diffusion patterns. It finds that online business primarily follows an innovation diffusion model, concentrating in developed eastern regions, while online shopping shows a more complex pattern, also benefiting peripheral areas, suggesting it has greater potential to bridge regional inequalities.
What Problem Does This Paper Address?
The rapid growth of e-commerce presents a significant economic and social phenomenon, particularly in China. While much research has explored the diffusion of internet technologies and online shopping, there's a notable gap in understanding the spatial diffusion patterns of online business itself, especially at a granular county level. Furthermore, existing studies often treat "e-commerce" as a monolithic concept, failing to distinguish between the business-side (online retail operations) and the consumer-side (online purchasing). This paper aims to address this by:
- Differentiating Online Business and Online Shopping: Recognizing these as distinct applications of e-commerce with potentially different diffusion mechanisms.
- Analyzing Spatial Diffusion at the County Level: Moving beyond broader regional or urban-rural divides to examine diffusion across China's diverse counties.
- Investigating Regional Inequality: Understanding how the spatial diffusion of e-commerce interacts with and potentially exacerbates or mitigates existing economic disparities between China's developed east and less developed west.
- Testing Diffusion Hypotheses: Evaluating the applicability of the "innovation diffusion hypothesis" (diffusion from centers of innovation) and the "efficiency hypothesis" (diffusion driven by economic benefits and overcoming distance friction) to both online business and online shopping.
Methods
The study employs a multi-faceted approach:
- Geographic Data Analysis: Using ArcGIS software to visualize and analyze the spatial distribution of online business and online shopping indexes. This includes:
- Exploratory Spatial Data Analysis (ESDA): To identify initial patterns and trends.
- Hot Spot Analysis (Getis-Ord Gi*): To pinpoint statistically significant clusters of high and low values for both online business and online shopping indexes, revealing areas of agglomeration.
- Core-Periphery Structure: Defining China's eastern region as the "core" and the western region as the "periphery" based on socioeconomic gradients to analyze diffusion from center to edge.
- Hypothesis Testing: The spatial diffusion patterns are analyzed to determine their alignment with the innovation diffusion hypothesis (gradient diffusion from core to periphery) and the efficiency hypothesis (diffusion driven by economic benefits, potentially in peripheral areas).
- Regression Analysis (OLS): Ordinary Least Squares regression is used to identify the key factors influencing the spatial diffusion of online business and online shopping. This helps to understand why the observed patterns emerge.
Datasets
The primary dataset used is the 2015 China County E-commerce Index, provided by the Ali Research Institute. This index is derived from e-commerce data from Taobao.com and includes:
- Online Retail Index (OBI): Measuring the level of e-tailing in each county.
- Online Shopping Index (OSI): Measuring the level of online shopping in each county.
These indexes are constructed using metrics like e-tailing/shopping density (number of retailers/consumers per population) and e-tailing/shopping trade/consumption levels.
Socioeconomic data for the regression analysis is sourced from the China County Statistical Yearbook and the Tianyancha.com platform. This includes data on:
- Offline retailers
- ICT service providers
- Logistics service providers
- Road infrastructure
- Per capita GDP
- Employment in the service industry
- Industrial output value
- Number of manufacturing enterprises
- Population density
- Per capita savings
Key Insights and Ideas
- Distinct Diffusion Pathways: The most striking insight is that online business and online shopping, while both forms of e-commerce, diffuse through China's counties via fundamentally different mechanisms.
- Online Business: The Innovator's Path: Online business diffusion is strongly tied to innovation centers and existing industrial infrastructure. It follows a clear gradient from the economically advanced east to the less developed west, aligning solely with the innovation diffusion hypothesis. This suggests that regions need a strong industrial base and technological capacity to foster online business.
- Online Shopping: A More Inclusive Spread: Online shopping, while also showing a core-periphery gradient, exhibits more complex patterns. It shows signs of "marginal reversal" or edge diffusion in some peripheral areas, suggesting it can overcome distance barriers and cater to local needs, aligning with both innovation diffusion and, in some cases, the efficiency hypothesis.
- Infrastructure vs. Economic Capability: The paper highlights that while basic internet infrastructure is crucial for both, online business development is heavily constrained by regional industrial endowments and economic factors. Online shopping, however, is more influenced by individual economic capabilities like income and savings, making it more accessible to a wider range of consumers, even in remote areas.
- E-commerce and Inequality: Contrary to the hope that online business might equalize regional disparities, the study suggests it has widened the gap. Online shopping, however, shows potential to reduce spatial consumption inequalities and improve welfare, not by equalizing economic output, but by increasing access to goods and services.
Key Results
- Online Business: Exhibits a clear gradient diffusion from east to west, with high values concentrated in eastern coastal regions. Hot spot analysis confirms this, showing significant clusters in the east.
- Online Shopping: Shows a wider distribution of high values, extending into some northern and northeastern regions, and even some remote western areas, deviating from a strict east-west gradient. Hot spot analysis reveals more dispersed high-value areas compared to online business.
- Hypothesis Alignment:
- Online business diffusion aligns primarily with the innovation diffusion hypothesis.
- Online shopping diffusion aligns with the innovation diffusion hypothesis overall, but also shows partial support for the efficiency hypothesis, particularly in peripheral areas.
- Regression Findings:
- Online Business: Positively influenced by offline retail, ICT service providers, high-grade roads, per capita GDP, service industry employment, and industrial output.
- Online Shopping: Positively influenced by per capita GDP, per capita savings, ICT service providers, logistics service providers, high-grade roads, and population density. The number of offline retailers also has a positive effect, which the authors interpret as online shopping often being a transformation from offline demand.
Limitations and Future Work
- Cross-Sectional Data: The study relies on cross-sectional data from 2015. The authors acknowledge that time-series data would provide a more dynamic understanding of diffusion patterns and that the current regression results are correlational rather than causal.
- Data Granularity: While county-level data is used, future work could benefit from even finer-grained data.
- Policy Implications: The authors suggest that while online business might not be a panacea for regional inequality, policies should still focus on improving infrastructure and support in marginal areas. They emphasize that online shopping offers a more promising avenue for bridging consumption inequalities and improving welfare in these regions.
- Future Research: The authors propose using panel data for more robust causal analysis and exploring the impact of e-commerce on the spatial structure of economies and regions, including whether it exacerbates or mitigates agglomeration advantages.
Summary updated