Size and sign of time savings

Abstract

INTRODUCTION
\n
\n
\nThe conventional approach in the U.K. has been to value all travel time changes at a constant
\nrate regardless of their size or direction. This ‘constant unit value’ approach was supported
\nby the 1980-86 UK DoT Value of Time Study (MVA/ITS/TSU, 1987). However, there has
\nalways remained a vocal body of opinion critical of this approach (see Welch and Williams,
\n1997, for references and discussion). Some of the main objections have been the following:
\n
\ni. small amounts of time are less useful than large amounts;
\n
\nii. small time savings (or losses) might not be noticed by travellers and any that are not
\nnoticed cannot be valued by those affected and so should not be valued by society;
\n
\niii. small time savings are said to often account for a large proportion of scheme benefits,
\nso that small errors in measurement might mean that the scheme is really of no benefit
\nto anyone;
\n
\niv. allowing small time savings to have ‘full’ value is said to inflate the measured total of
\nbenefits and so lead to schemes (often road schemes) being wrongly found to have
\nsufficient net benefit to justify implementation;
\n
\nv. time savings are less highly valued than are time losses, according to surveys, and so
\nshould have a lower unit value when evaluating schemes.
\n
\nBoth aspects relate to the possible non-constancy of the value of time for a given journey
\nmade for a given purpose (clearly, it is much less controversial, and indeed standard practice,
\nto allow for variation by purpose and traveller type).
\n
\nThe practical difficulties are twofold. On the one hand, it is difficult to overcome the lay
\nreaction that small time savings have little or no value, as well as the feeling that losses are
\nmore important than gains. On the other hand, if these points have any empirical relevance,
\nthey cause major problems for the cost-benefit calculus, as losses and gains will not cancel
\nout, and time savings cannot be directly aggregated.
\n
\n
\nAlthough they do not recommend that values differentiated by size and sign should be used
\nfor appraisal, the HCG/Accent (1999) Report (AHCG) notes that [p 259]
\n
\n"For any level of variation around the original journey time, gains (savings) are valued
\nless than losses. For non-work related journeys, a time savings of five minutes has
\nnegligible value."
\n
\nA recent paper by Gunn (2001) notes that corroborative results are available from a reanalysis
\nof the 1988 Dutch value of Time study.
\n
\nFor reasons which will be carefully rehearsed in this paper, we do not believe that the
\nconclusion on the differences between gains and losses is safe. This is based on an extensive re-analysis of the AHCG data. We have found it harder to reach a conclusion on the issue of small time savings, we agree with AHCG that their data undoubtedly implies a lower valuation: we have some concerns, nonetheless, as to the interpretation which should be placed on this.