ABSTRACT
The theoretical ideas presented in this subsection first appeared in Wilson (1976-A) and were subsequently developed (Wilson and Kirkby, 1980) to revise the idea of the nature of the splitting factor in this particular model. Originally, the purpose of this model-building exercise was to teach the author something about catastrophe theory. It turns out, in this case, not so much to illustrate the notion of jumps, but that of hysteresis, which is potentially important in modal choice.
