世界银行-道路安全背景下统计生命的价值_一种新的价值转移方法（英）-2025_9页_1mb

Summary of "The Value of a Statistical Life in the Context of Road Safety: A New Value Transfer Approach"

Core Content

This paper introduces a new value transfer approach to estimate the value of a statistical life (VSL) in the context of road safety, particularly for low- and middle-income countries (LMICs) that lack country-specific VSL estimates. The VSL is a critical input for cost-benefit analysis (CBA) and the socio-economic cost calculations of road crashes.

Main Objectives

• To develop a new VSL transfer function for road safety, based on recent data and income elasticities.
• To provide guidance for using these functions in CBA and road crash costing in countries without official VSL estimates.
• To highlight the need for more VSL studies in LMICs and the importance of developing transfer functions for non-fatal injuries.

Key Findings

• VSL is strongly correlated with income per capita, which is consistent with the theoretical assumption that safety is a "normal good."

• Two base VSLs are established:

  • LMIC: 0.404 million USD
  • HIC: 3.206 million USD

• The VSL transfer functions are:

  • LMIC: $\mathrm{VSL} = 0.404^{*}(\mathrm{Y} / 5,726)^{1.2}$
  • HIC: $\mathrm{VSL} = 3.206^{*}(\mathrm{Y} / 42,087)^{0.8}$
  • Where Y is the Gross National Income (GNI) per capita in USD (2020 prices), and VSL is in million USD.

• VSL ranges:

  • LMIC: 22,000 USD to 1.1 million USD
  • HIC: 1.2 million USD to 4.8 million USD

• Income elasticities:

  • LMIC: 1.2
  • HIC: 0.8

• These elasticities are used to adjust VSL estimates based on a country's income level, ensuring the relevance of the transfer function to the local context.

Methodology

• A unit value transfer approach is adopted, using GNI per capita as the income indicator.
• Data sources include:
  • Official government VSLs from 7 countries
  • Data from the SafetyCube project for 17 HICs
  • Academic literature from Scopus, yielding 142 publications
• The dataset includes VSLs from 32 countries, with 25 HICs and 7 LMICs.
• Outliers from Ethiopia and Egypt were excluded due to unusually high VSL/GNI ratios.
• Sensitivity analysis ranges are recommended:
  • LMIC: 1.0–1.4
  • HIC: 0.6–1.0

Results and Implications

• The new VSL transfer function yields lower VSL ratios for LMICs compared to the iRAP rule of thumb at lower income levels, but higher VSLs at higher incomes.
• The OECD base value (3.9 million USD) results in lower VSLs for LMICs due to the larger income interval used in the transfer.
• The VSL to income ratio:
  • Low-income: 46
  • Lower-middle-income: 58
  • Upper-middle-income: 77
  • High-income: 76
• The VSL accounts for over 90% of the total value of a prevented fatality in most countries, with an unweighted average of 94%.

Recommendations

• The new VSL transfer function should be used for CBA and road crash costing in countries lacking official VSL estimates.
• More VSL studies are needed, especially in LMICs, to improve the accuracy of future transfers.
• Transfer functions for non-fatal injuries should be developed, as they are also essential for CBA.

Conclusion

The paper presents a robust and updated VSL transfer method that accounts for income differences and aligns with current CBA practices. It emphasizes the importance of using GNI over GDP for consistency with World Bank income classifications, and highlights the need for more local VSL research in LMICs to enhance the reliability of cost-benefit analyses in road safety.