Option value estimates for rural bus services in England

Abstract

OPTION VALUE ESTIMATES FOR RURAL BUS SERVICES IN ENGLAND. The topic of Public Transport Option Values is bedevilled by a shortage of empirical evidence. This state has arisen due to many regarding the concept as spurious. They feel that a person’s Willingness To Pay for the continued provision of a Public Transport service that person has no plan to use would be discovered as a large Consumer Surplus on those few days on which unplanned use was made of the service. Yet the UK Department for Transport has decided, after much consideration, that such values can be included in appraisals in principle, and have published official values on its WEBTAG online guidance site. The paper will demonstrate how that advice has influenced the official advice in several other countries. The purpose of the present paper is to report on emerging Option Values estimates from a current survey, in Shropshire, England, as part of a larger project on ‘Buses and the Economy’ funded by the UK Department for Transport and Greener Journeys. The context is one where the availability of public money for subsidies to public services is decreasing, and it is therefore necessary to establish the costs to society of losing accessibility from small outlying settlements to the ‘county town’, in this case the city of Shrewsbury. Naturally, the paper takes the opportunity to compare the values emerging from our new survey with past evidence, with a view to informing any future updating of the official values. Additionally, the paper will revisit the justification for including such values in the first place. Central to that justification will be the near impossibility of spotting rare extremely high values of Consumer Surplus in Willingness To Pay (WTP) surveys, both because of their rarity and because no Revealed Preference experiment could contain the desired trade-offs, and because Stated Preference (SP) surveys are unsuitable for the purpose. There are two reasons why SP surveys are not suitable. Firstly, it would impose a distortion on the survey design to give a range of trade-offs capable of accurately estimating such high WTP values on just that survey day – it being much more likely that the respondent would answer for a normal day, when the all presented alternatives were actually available. More likely still, the respondent will choose not to respond to that question, believing the survey to be aimed at regular Public transport travellers. Furthermore, the respondents might anyway be excluded from the analysis if it were realised by the analyst that the non-chosen mode was not available that day. Having demonstrated the need for values, on a range of grounds including those just given, the paper will then detail how the Shropshire experiment was designed and mounted. The need for that arises from the rareness of studies in this area. While the present study owes a lot to previous work in this area, including via an overlap of authors, there is certainly no existing blueprint for how to proceed. It is to be hoped that future studies will avoid pitfalls, and benefit more generally, from a detailed exposition of the steps we have made over the last several months. In particular, attention has been paid both to the appropriate ‘Payment Mechanism’ and the ‘non-transport alternative’. In

England, the obvious Payment Mechanism is the Council Tax, which is a local tax allowing Local Authorities to top up their allowance from central government to support local services. Although well known to most respondents, there are various discounts and exemptions that we have needed to stress do not apply to the Council Tax changes we present. The perceived need for the inclusion of a non-transport alternative, Library opening hours, is to both distract from the main purpose of the survey, and to help calibrate our survey findings against existing surveys values for that non-transport alternative. The Shropshire survey itself is currently running, but the early results (from what we would have regarded as our pilot had any significant changes been required) have shown an excellent standard of completion. As part of that, we have checked that there is a good range of responses to our Stated Preference experiment. Individuals have been willing to switch between alternatives as their attributes have been varied, and different individuals have responded differently. This excellent standard of responses is no doubt due to our decision to insist on face-to-face interviews, by trained and experienced interviewers, in the respondents’ own homes. The final sample contains 200 such completed interviews. Past studies have not always gone so well, the paper will discuss the likely reasons for that. The paper will end by summarising our results, and giving advice for future official forecasts.