Using meta-regression to determine Noise Depreciation Indices for Asian airports

Abstract

The external costs of aviation noise are an important input in policy assessment for cost-benefits analysis. The Noise Depreciation Index (NDI) is used to capture the externality costs through measuring the depreciation of property prices (PPs) exposed to aviation noise. This paper summarizes existing studies on NDI and examines the underlying differences in order to transfer these NDI values to other parts of the world, where NDI estimates are not directly available. We find that higher wealth, expressed in terms of PPs, relative PPs or income, result in higher values of NDI. This means that wealthier households de-value the PPs more than the average in the presence of aircraft noise. The income dependence allows the NDI estimates to be transferred to other locations using local PPs or income for cost–benefit analysis. Estimates of NDI for some Asian countries using the meta-regression results are also provided.

Keywords:

• Noise Depreciation Index
• hedonic price method
• meta-analysis
• airport noise costs

Acknowledgements

The study was conducted while the author was at the University of Cambridge, UK and was partly funded by OMEGA. Any opinions, findings, conclusions, or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of OMEGA.

Notes

1. For meta-analysis, we have picked one estimate for one location from one study. In Table 2, however, all the relevant estimates for one location are included to increase the sample size.

2. Cohen and Coughlin (2007) multiply their NDI estimate with a spatial multiplier to account for the effect that if property prices around a neighborhood increases, the prices of a specific property in that neighborhood will automatically increase. This results in an NDI of 7.2% per dB, which is significantly larger than other NDI estimates. Noise, however, affects all properties in the neighborhood and since the effect is already captured in the HP estimates, the multiplier effect is not applicable. See Small and Steimitz (2006) for further discussion on this.