Consensus formation: The case of using cell phones while driving

Several models (including the widely used Sznajd model) have been proposed in order to describe the social phenomenon of consensus formation. The objective of the present paper is to supplement the simulations based on these models with a ``real world simulation''; it considers a situation which can be considered as an ideal laboratory for analyzing consensus formation, namely the authorization or prohibition of using cell phones while driving. This is a convenient laboratory for several reasons (i) The issue was raised in similar terms in all industrialized countries, a circumstance that facilitates comparative analysis by providing a {\it set of observations} (as opposed to single observations generated, for instance, in election contests). (ii) This is a situation where we happen to know the rule a ``rational'' agent should follow. (iii) Because the issue is a matter of life and death, the phenomenon can be considered as fairly robust with respect to various, endogenous or exogenous, sources of noise. (iv) The relevant data are available on the Internet and in newspaper data bases. \qL Our observations strongly suggest that there is a missing variable in most consensus formation simulations. In its conclusion, the paper calls for a large scale effort for identifying, documenting and analyzing other real-world cases of consensus formation.