2007年-世界发展银行全球_Domestic_Trade_Impacts_of_the_Expansion_of_the_National_Expressway_Network_in_China_28页_1mb

Summary of Domestic Trade Impacts of the Expansion of the National Expressway Network in China

Core Content

This report examines the domestic trade impacts of the expansion of China's National Expressway Network (NEN), focusing on how improved road infrastructure has influenced trade flows, economic integration, and regional development. It uses a spatial analysis framework to estimate the effects of expressway construction on trade and provides a policy simulation based on these estimates.

Main Objectives

• To quantify the impact of expressway network expansion on domestic trade.
• To provide a spatially explicit analysis of trade flows between cities and provinces.
• To estimate trade expansion and travel time savings resulting from the improved road network.
• To outline methodological approaches and suggest improvements and extensions for future studies.

Key Findings

• Infrastructure Development: Between 1990 and 2005, China constructed nearly 41,000 km of high-grade tolled expressways, forming the National Trunk Highway System (NTHS), which was later renamed the National Expressway Network (NEN). This was part of a larger effort to expand the road network, with an estimated 400,000 km of local and township roads also improved.

• Economic Growth: The NEN was designed to connect major cities and improve economic and social integration, especially in poorer regions and interior parts of the country. The network is structured around 7-9-18 corridors, combining radial and grid patterns to maximize coverage and connectivity.

• Trade Expansion: The report estimates that the expansion of the NEN led to a 120.6% increase in trade flows between cities, from 4,814 to 10,619 billion Yuan. This is attributed to reduced travel times and enhanced transport connectivity.

• Methodology: A gravity model is used to estimate trade flows, where trade volume is proportional to economic size and inversely proportional to distance or travel time. The model is spatially explicit, using GIS data to compute travel times and allocate flows to specific network links.

• Policy Simulation: The analysis predicts trade flows under a scenario with improved travel times due to the NEN. It holds GDP levels constant at 1997 levels to isolate the impact of transport improvements.

• Limitations: The data on internal trade flows is incomplete, and the results are likely a lower bound. The report acknowledges that economic growth itself has a significant impact on trade, and infrastructure improvements may not be the only factor.

Methodology Overview

• The gravity model is estimated using lognormal regression with imputed trade flows, province GDP, and baseline travel times.
• Travel time is calculated using GIS data, including design speeds for different road types.
• Data sources include:
  • GIS road network data (from ACASIAN).
  • City and province GDP (adjusted to 2005 levels).
  • Inter-provincial trade flows derived from input-output (IO) tables.
• The methodology includes:
  • A spatial interaction model to allocate flows to specific network links.
  • Log transformation and exponential prediction to estimate trade flows.
  • Additive correction factor applied to account for model uncertainty.

Data Set Construction

• GIS road network data includes 20,899 line segments in the base network and 31,538 segments after including expressways.
• City data is based on population thresholds (500,000 or more), with GDP allocated based on population share.
• Trade flows are estimated using IO tables from 1992, 1997, and 2002. Imputation methods are used for missing data, particularly for smaller provinces like Hainan and Tibet.

Network Modeling and Estimations

• Fastest routes are calculated using the Dijkstra algorithm.
• Travel time savings are significant:
  • Average travel time drops from 38.2 hours to 28.6 hours (a 25% reduction).
  • Median travel time reduces by 8.5 hours.
  • Maximum travel time decreases from 104.0 hours to 76.2 hours.
• Gravity model results:
  • Equation:
    $$
    \ln \hat{T}{ij} = -5.128 + 0.741 \cdot \ln E_i + 0.8735 \cdot \ln M_j - 2.016 \cdot \ln d{ij}
    $$
  • The model has an adjusted $R^2$ of 0.42, which is acceptable but lower than international trade studies.

Possible Improvements and Extensions

• Intra-provincial trade and international trade could be incorporated for a more comprehensive analysis.
• More accurate trade data would improve the robustness of the model.
• Spatial allocation models could be enhanced to better reflect domestic trade patterns.
• A multi-modal transport model integrating rail and waterway data could be developed, though mode switching at intersections is a complex issue.
• The friction parameter (representing travel time and distance) is more influential than the economic size parameters in the gravity model.

Conclusion

The expansion of China's National Expressway Network has had a significant impact on domestic trade, reducing travel times and increasing interprovincial trade. The spatial analysis approach provides a valuable framework for assessing the economic benefits of infrastructure development. However, due to data limitations, the results should be interpreted as lower bound estimates, and further research is needed to refine the methodology and include additional trade dimensions.