Willingness-to-pay for noise abatement in Singapore

Abstract

Noise in urban centres adversely affects the health of residents and the liveability of the city. To inform the formulation of noise control policies, the marginal benefits of publicly-provided noise abatement are examined in this thesis. The studies in this thesis were conducted in Singapore, a dense urban centre with elevated levels of noise. Marginal benefits of reducing noise from construction activity and roads were estimated as these are key contributors to noise pollution in Singapore. Stated preference surveys have been used to estimate the marginal benefits of noise abatement. The veracity of estimates obtained from these surveys depends on respondents' understanding of the information provided in the survey questionnaire. Since noise is measured in decibels, which is scaled logarithmically, respondents may not understand changes to the noise levels when described in physical units. In response to the challenge of communicating different noise levels to respondents, a wide range of textual descriptions has been developed and utilised. However, all these text-based descriptions require respondents to read, understand, and interpret the descriptions. Respondents' interpretation of these textual descriptions may be subjective, leading to inaccurate estimates of their willingness-to-pay for noise abatement. The surveys conducted in this thesis seek to overcome the challenges of describing construction and road noise textually with an innovative approach of describing changes to noise levels with actual audio recordings. Use of noise recordings enabled survey respondents to hear the change in noise levels without having to subjectively interpret descriptions of noise, hence removing the ambiguity of text-based descriptions. To test the effect of audio- and text-based representations on willingness-to-pay for noise abatement, a text-based survey was designed. With the exception of the description of noise, the valuation questions in both survey questionnaires are the same, enabling the isolation of the effect of noise representation on willingness-to-pay. Respondents to the text-based survey were found to have a lower willingness-to-pay on average. Further analysis suggests that this difference was due to a lack of understanding of the information in the survey questionnaire. This result suggests that previous text-based questionnaires may have underestimated the willingness-to-pay for noise abatement. Finally, three sources of heterogeneity were examined, namely prior private spending on noise abatement measures, hearing sensitivity, and distance between noise sources and receptors. These sources of heterogeneity were found to influence the willingness-to-pay for noise abatement, suggesting that cost-benefit analysis of publicly-provided noise abatement should take into account these factors when estimating the marginal benefits of noise abatement. The estimated marginal benefits can inform cost-benefit analysis of noise abatement policies. In the absence of government policies, market failures cause noise emissions to be higher than optimal. However, without rigorous cost-benefit analysis, governments may under- or over-provide noise abatement and these policies may not improve the welfare of society. Hence, cost-benefit analysis can inform the design of noise abatement policies and ensure that policy-makers know whether a policy results in a potential Pareto improvement. Show more