By Jonathan Wolff – Blavatnik School of Government, University of Oxford
Jones-Lee and Aven’s paper is a very helpful account of the quantitative cost-benefit approach to safety decision-making as it has developed in the USA and Europe, and of which Jones-Lee has himself been the leading figure in the UK. The paper is a very fair representation of the philosophical and economic rationale for the approach, the methods available for eliciting public values, and the difficulties of these methods. Indeed many of the most serious difficulties were recorded in Jones-Lee’s own work as long ago as Jones-Lee et al., 1985, and he has worked tirelessly since to find ways to meet the challenges, as demonstrated in this paper.
Jones-Lee and Aven conclude that formal methods, based on willingness-to-pay measures, should only be part of a decision-support process rather than the sole determinant of the values to be used in public safety decision-making. Partly their reason for this seems to be the uncertainty that any derived figures may have, but their proposal seems not only sensible, but, in fact, how transport safety decisions are made in the UK at the present time. I would go somewhat further and say that the UK operates by an implicit triangulation of political judgement, decision analysis by means of formal methods, and public opinion. Probably, many other countries do so too.
To explain, in an appendix to Wolff and Orr (2009) we looked at the question of how, in fact, VSL figures in the UK had been derived, after the abandonment of “human capital” approaches in favour of taking into account public preferences for safety. As far as we were able to tell, the first main move was made in a paper by M. Q. Dalvi for the Department of Transport (1988). Dalvi reviewed a number of US studies as well as Jones-Lee et al. (1985) and derived a VPF figure that was partly based on Jones-Lee’s calculation. However, Jones-Lee’s recommended figure was reduced first by taking the (lower) median valuation rather than the then recommended mean, and then, merely at the stroke of a pen, reducing the resulting figure from £890,000 to £500,000 on the claimed justification of the government’s preference for speedy movement of traffic (p. 34). No further explanation of the derivation of that valuation was given.
This figure of £500,000 in 1988 prices is equivalent to around £1.25m in 2017. Currently a figure of closer to £1.8m is used (RSSB 2016), but as I understand it, this also includes around £650,000 for lost output (Dept of Transport ND), so the figures in use today remain remarkably close to the original, hybrid, judgement. A number of new studies have taken place over the years, many by Jones-Lee and his associates, but tend to recommend a range rather than a point value. This gives decision-makers some discretion to follow their judgement but at the same time claim support from rigorous studies. Typically the range includes the currently used figure.
While I have been, at times, sceptical of the accuracy of some of the formal measures used to provide a precise valuation of a statistical life, I take some comfort in a political feedback system in which, should government decisions about safety seem to lax or too strict (“health and safety gone mad”) the public is not short of ways of letting the government know about this, as part of the process. Local newspapers and campaigns will be one main way in which this message is conveyed. Hence, for as long as there is no virulent public outrage at government safety policy, we can cautiously assume a level of tacit public approval, in addition to political judgement and formal derivation.
Before closing, however, I would like to underline Jones-Lee and Aven’s point that although consistency is to be valued, this is not an area where one size fits all. Two contrasts are important: how much can an individual control their exposure to risk; and how are safety improvements funded? In some areas, such as train safety, clearly a given individual has virtually no say in the level of risk they take if they are to travel. But in road safety matters differ, both in control, and in payment, and we need to take into account consumer response. For example, if car safety tests become very expensive, some people will break the law, magnifying their risk, but saving their money. In contrast, wealthier people might decide to spend more on their own safety by purchasing safer cars or other devices. Although, of course, for reasons of third party effects this is an area fraught with difficulty, regulating private purchases where risk is to some degree an individual matter is not exactly the same topic as taxing people to pay for a single level of safety. Hence there can be room for further variation, just as their should be for “dread risk”, which is nothing other than respecting the preferences of the public, if it is shown that it does really exist.
References
Dalvi, M.Q., (1988) The Value of Life and Safety: A Search For A Consensus Estimate, Department of Transport.
Dept of Transport (ND) https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/254720/rrcgb-valuation-methodology.pdf (Viewed December 28th, 2016).
Jones-Lee M., Hammerton M., Philips P.R. (1985) ‘The value of safety: results of a national sample survey‘ Economic Journal 95(377) 49-72.
RSSB, Taking Safe Decisions – Safety-Related CBA https://www.rssb.co.uk/risk-analysis-and-safety-reporting/risk-analysis/taking-safe-decisions/taking-safe-decisions-safety-related-cba (Viewed December 28th, 2016)
Wolff J, Orr S. (2009) Cross-sector weighting and valuing of QALYs and VPFs. A Report for the Interdepartmental Group for the Valuation of Life and Health. Final Report 2009. https://www.ucl.ac.uk/health-humanities/docs/IGVLH.pdf (Viewed December 28th 2016)