Travel surveys are subject to a wide range of errors, including sampling error, non-response error and measurement error. This paper considers the issue of measurement error in the estimation of the amount of travel on 'employer's business' by personal vehicle. 'Employer's business' travel is defined as travel on the business of the employer (for example, trips to meetings, trips between work sites, and so forth), but excludes travel between home and work (commuting). It uses the example of the Survey of Motor Vehicle Usage conducted periodically by the Australian Bureau of Statistics. It shows that, by comparison with a number of other travel surveys, the Survey of Motor Vehicle Usage considerably over-estimates the amount of travel on 'employer's business' performed in private motor vehicles. Given that 'employer's business' travel is generally accorded a higher value of travel time savings in economic evaluations of road improvements, such an error could have a substantial impact on the viability of various road improvements. A reason for the difference is proposed, based on the wording of the specific question and the type of respondent completing the survey, which leads to a significant source of measurement error. (a).
Abstract