Existing noise planning and regulation guidelines in the United Kingdom attempt to define the proportion of a community likely to be annoyed at a particular noise level. But these guidelines are not based on any systematic comparison of the effects of different types and levels of noise, Ian Diamond told a CEPR lunchtime meeting on 2 June. Ian Diamond is a CEPR Research Fellow and a Lecturer in Social Statistics at the University of Southampton.
Diamond began by arguing that some aspects of the financial costs of noise control could be measured straightforwardly, such as the cost of a noise barrier or of extending a by-pass in order to reduce traffic noise in a local community. The financial costs to individuals were harder to measure. Though there is considerable evidence that noisy environments depress housing prices, how far an individual is prepared to forgo peace and quiet for the sake of purchasing a cheaper house is hard to assess. The problem of `costing' noise becomes even harder when non-monetary factors are involved: there is no systematic evidence, for example, on the links between noise and ill health.
Most previous work on the costs of noise had focused on the relationship between the volume of noise and the level of annoyance it caused, but none had succeeded in establishing a close connection. The main reason for this, Diamond observed, was that an individual's reaction to a particular noise is a function of many factors, both acoustic and non-acoustic. As an example, in a recent survey of reaction to aircraft around small aerodromes conducted with John Ollerhead, Diamond had found that the level of noise was less important in determining the level of annoyance than non-acoustic factors such as the perceived importance of particular types of flying, the care and concern that the aerodrome in question showed for the local community, and the form of housing tenure in the area.
This suggested the need for a more sophisticated model of the relationship between noise and annoyance, which would incorporate both acoustic and non-acoustic factors. No such model could be applied uniformly, so it would have to be supplemented by a `Decision Support System'. This would be composed of information concerning both the neighbourhood (whether or not the neighbourhood is residential, the type of housing, and expectations about noise levels) and features of the noise source (such as its level, intermittency, impulsiveness and tonality)