Notes on the interpretation of traffic accident data and of risk homeostasis theory: a reply to L. Evans

Risk Anal. 1986 Mar;6(1):95-107. doi: 10.1111/j.1539-6924.1986.tb00197.x.


A quick and summary comment is made on a paper by Evans, who set out to "definitively answer" the question "whether there is any validity" in risk homeostasis theory (RHT), and who arrives at the conclusion that the "theory should be rejected" because the data "provide the clearest evidence refuting the theory." It will be argued here that the seven cases Evans puts forward against RHT do not justify such a rather boldly-worded conclusion. On the contrary, the data presented by Evans can be seen to be either not inconsistent with RHT (Case 1), in error (Case 2), or fail to take account of relevant economic changes (Case 3) and of pertinent exposure variables (Case 5), incomplete or the product of guess-work (Case 4), irrelevant (Case 6), nonrepresentative (Case 7) and in all cases open to alternative interpretations that do not conflict with the theory under debate. Moreover, Evans' reasoning seems to betray not only careless reading of what RHT does in fact say, but also a fundamental misunderstanding of the notion of target risk in particular. Refuting a "refutation" does not, of course, imply that RHT is valid, or valid at least within as yet to be defined limits. Attempts to address the validity question should preferably take the form of well-controlled field experiments, instead of retrospective analyses of multi-interpretable archival data that can be debated ad infinitum.

MeSH terms

  • Accidents, Traffic*
  • Humans
  • Models, Theoretical*
  • Risk