top of page

17 NOV 2010 (WED) | 19:00 - 20:00




Chief Research Political Scientist, Institute of Transport Economics, Norway




19:00 - 20:00




Institute of Transport Studies, The University of Hong Kong


Rationality is a well-established ideal for transport safety policy. Indeed, it has been claimed that if we only are rational enough, we can solve all our problems. This, however, can only be the case if rationality is a uniquely defined ideal, i.e. if it is always clear that only one solution to a problem is rationally justified and no other solution can be claimed to be rational. A paradox of rationality refers to any situation in which conflicting choices can both be defended as rational. When a paradox of rationality occurs, the normative guidance provided by the principle of rationality normally breaks down. This paper will first outline what the implication for road safety policy of an ideal of perfect rationality. Based on this ideal, a set of paradoxes of rationality will be discussed. More specifically, the following paradoxes will be discussed: (1) Choices between policy options that produce identical safety benefits may be arbitrarily influenced by irrelevant background characteristics, i.e. violate the axiom of independence of irrelevant alternatives. (2) Choices between policy options based on aggregate willingness-to-pay for safety improvements may violate individual preferences, i.e. there is an unintended reversal of preferences resulting from the aggregation of preferences. (3) Choices between policy options based on aggregate willingness-to pay for safety improvements may disregard the intensity of individual preferences, i.e. favour those who have a weak preference for safety at the expense of those who have a strong preference for safety. (4) Choices between a policy option favouring the poor and a policy option favouring the rich may favour the rich, because the declining marginal utility of money makes it impossible for the poor to compensate the rich in utility terms, i.e. the threshold a benefit-cost ratio satisfying the criterion of a potential Pareto improvement may be higher for the poor than for the rich. (5) A policy option that looks attractive and passes the benefit-cost test ex ante may fail an ex-post compensation test if utility depends both on wealth and health and is more risk aversive with respect to health than with respect to wealth. (6) Some road safety measures will be cost-effective from a societal point of view, but not necessarily from an individual point of view, i.e. there is sometimes a conflict between individual and collective rationality. (7) Resource allocation based on grants from the central government stimulates strategic coalition building among regional and local units of government that undermines an efficient allocation of resources. The implications of these paradoxes of rationality for road safety policy will be discussed.


Rune Elvik is a political scientist from the University of Oslo. He attained the degree of doctor of political science (dr. polit) in 1993 and the degree of doctor of philosophy (dr. philos) in 1999. In 2007, he attained the ph D degree at Aalborg University in Denmark. His main areas of research include evaluation of the effects of road safety measures, research synthesis by means of meta-analysis, and cost-benefit analysis. Since 1994, he has been chief research officer in charge of risk analyses and cost-benefit analyses. This area had a strategic research programme on meta-analysis from 2000 to 2008. Rune Elvik has taken part in several international research projects organised by the European Commission, the OECD, the European Transport Safety Council and the US Transportation Research Board. During the years 1997-2004 he was associate editor of Accident Analysis and Prevention. From 2005, he has been one of the editors-in-chief of the journal. From 1999, Elvik has been a member of the committee for safety data, analysis and evaluation of the Transportation Research Board. From 2008 to 2010, Elvik was professor of road safety studies at Aalborg university in Denmark (20 % working time). From 2009, he is professor of road safety studies at Lund university in Sweden (20 % working time).

8 views0 comments

Recent Posts

See All


bottom of page