2022-05-19-IMF-Road_Quality_and_Mean_Speed_Score_34页_1mb

Road Quality and Mean Speed Score Summary

Core Content

This working paper introduces a novel measure of cross-country road quality based on the travel mean speed between large cities, using data from Google Maps. The proposed measure, the Mean Speed (MS) score, serves as a proxy for road infrastructure quality and access, offering an alternative to traditional indicators like the World Bank's Rural Access Index (RAI) and the World Economic Forum's Quality of Road Infrastructure (QRI) score.

The MS score is defined as the harmonic mean of travel speeds between the largest city and other significant cities in a country, calculated by dividing the total distance by the total travel time. The paper also presents two additional variations: the geometric mean speed (gMS) score and the adjusted mean speed (aMS) score, which accounts for geographic obstacles.

Main Points

• MS Score: A robust and easily estimable proxy for road quality and access, based on travel speeds between major cities.
• gMS Score: A geometric mean of travel speeds, which penalizes outliers and provides a different perspective on speed distribution.
• aMS Score: An adjusted version of the MS score that accounts for geographic terrain, making it a more accurate representation of road quality under ideal conditions.

The MS score is found to be highly correlated with existing road quality indicators, such as RAI and QRI, and offers a more frequent and cost-effective alternative for data collection and analysis.

Key Information

• Range of MS Scores: The MS scores range from 38 km/h (23.6 mph) to 107 km/h (66.5 mph) across over 160 countries.
• Correlation with Traditional Indicators: The MS score is strongly correlated with RAI (ρ = 0.56) and QRI, making it a reliable complement.
• Data Sources: The paper uses the Google Maps API to estimate travel times and distances. It also incorporates data from the World Bank, World Economic Forum, and the Legatum Prosperity Index to validate the MS score.
• Geographic Adjustments: The aMS score adjusts for terrain by using the ratio of actual road distance to straight-line (crow-flies) distance, providing a more accurate measure of road quality.
• Sample Countries: The dataset includes 760 cities in 162 countries, with a minimum of three cities per country. Countries with fewer than two cities over 80 km from the largest city are excluded.

Methodology Overview

• City Selection: Major cities are identified using UN population data. Each country is required to have at least three cities, and only cities more than 80 km from the largest city are included to avoid biased results from short-distance travel.
• Speed Calculation: The MS score is calculated as the harmonic mean speed, while the gMS score is the geometric mean. The aMS score is derived by adjusting travel time for terrain using the crow-flies ratio.
• Validation: The MS score is validated against traditional indicators such as GDP per capita, road density, RAI, and QRI. The results show that MS, gMS, and aMS scores are highly correlated (above 0.96), with MS being preferred for its straightforward economic interpretation.

Applications

• Welfare Calculations: The MS score can be used in economic models to assess the impact of road quality on welfare.
• Public Investment Management: It helps in the planning and prioritization of infrastructure investments by providing a clear and actionable metric for road quality and access.
• Policy Analysis: The MS score is a useful tool for policy makers to evaluate and compare road infrastructure across countries, especially in regions with limited data availability.

Conclusion

The MS score is a reliable, cost-effective, and easily replicable measure of road quality and access. It complements traditional indicators and provides a more accurate and frequent alternative for assessing the efficiency of road networks in moving people and goods. The aMS score, which adjusts for geographic terrain, offers a theoretical construct of road quality under ideal conditions, further enhancing the accuracy of the measure.