2015年-世界发展银行全球_Transport_Infrastructure_and_Welfare___An_Application_to_Nigeria_67页_3mb

Transport Infrastructure and Welfare: An Application to Nigeria

Core Content

This paper investigates the impact of transport infrastructure on welfare in Nigeria, using a comprehensive dataset and a novel instrumental variable (IV) approach to address endogeneity issues. The study is part of a broader effort by the World Bank to inform development policy through empirical research.

Main Objectives

• To assess the differential development impacts of alternative road construction.
• To prioritize road programs based on their expected development outcomes.
• To evaluate how reducing transport costs affects various welfare indicators in rural Nigeria.

Key Findings

• Transport cost reduction leads to significant improvements in several welfare indicators:

  • Crop revenue increases by approximately 6.4%.
  • Livestock revenue increases by 3.4%, though this result is less robust.
  • Non-agricultural income rises by 3.3%.
  • All-year employment increases by 0.4% (male) and 0.3% (female).
  • Multi-dimensional poverty index (MPI) decreases by 2.6%.
  • Wealth index improves by 2.0%.
  • Local GDP increases by 5.0%.

• The natural path IV, which represents the time it would take to walk to the nearest market without roads, is used to address the endogeneity of road placement. This IV is more accurate than traditional straight-line instruments in capturing the most cost-effective route.

• The study also considers spatial sorting of households and endogeneity of market locations, using marketshed fixed effects and carefully selected control variables to mitigate potential biases.

• The elasticities of welfare indicators to a 10% reduction in transport costs are summarized in Table 1, showing varying degrees of impact across different metrics.

Methodology

• The paper uses two household surveys (LSMS-ISA and NDHS) and a raster dataset on local GDP to measure welfare outcomes.
• A highway development model (HDM-4) and GIS tools are employed to estimate actual travel costs to market.
• The natural path is derived from the topography and terrain, and is used as an IV to estimate the causal effect of transport costs on welfare.
• The empirical framework includes both seemingly unrelated regression (SUR) and three-stage least squares (3SLS) methods to account for endogeneity.

Robustness Checks

• The study performs robustness checks using Conley Bounds to assess the sensitivity of the estimates to different degrees of exogeneity.
• Euclidean distance as an IV yields very similar results to the natural path.
• The results are robust to the inclusion or exclusion of various control variables, including land, labor, and credit.

Policy Implications

• The findings highlight the importance of transport infrastructure in promoting economic growth and poverty reduction.
• They suggest that prioritizing road construction in areas where the impact on welfare is highest can lead to more efficient use of scarce development resources.
• The study underscores the need for selective investment in transport infrastructure to maximize benefits for households and regions.

Conclusion

• The paper provides empirical evidence that reducing transport costs in Nigeria has a positive impact on multiple dimensions of welfare.
• It demonstrates how instrumental variables and geospatial analysis can be used to improve the accuracy of welfare impact assessments in the context of transport infrastructure.
• The results are robust to various alternative specifications and endogeneity corrections, offering a policy-relevant framework for evaluating transport projects in Sub-Saharan Africa.

Key Variables and Data Sources

• Natural path (IV): Time to walk to the nearest market without roads.
• Transport cost to market: Estimated using HDM-4 and GIS tools.
• Welfare indicators:
  • Crop revenue
  • Livestock revenue
  • Non-agricultural income
  • All-year employment
  • Multi-dimensional poverty index (MPI)
  • Wealth index
  • Local GDP
• Data sources:
  • 2010 LSMS-ISA (Nigeria)
  • 2008 NDHS (Nigeria)
  • Ghosh et al (2010) nighttime lights raster data set

Methodological Innovations

• Introduction of the natural path as an IV to capture the most cost-effective route for road construction.
• Use of geospatial data to estimate local GDP and transport costs at a high resolution.
• Application of 3SLS and SUR to account for endogeneity in transport cost data.

Related Literature

• The paper contributes to a growing body of literature on the effects of transport infrastructure on economic well-being.
• It builds on previous studies that have examined the impact of transport on GDP, employment, and poverty, but highlights the need for better identification strategies to address reverse causality and spatial sorting.

Summary of Elasticities