The first of the two modelling problems is the older one: spatial interaction models of a now almost standard form have been available to describe consumers' behaviour since the early-1960s. The second problem is newer and is concerned with the evolution of structure. The usual implicit hypothesis about consumers behaviour is that there is a trade-off between the benefits of centre size (cheaper prices, more choice and so on) and the increasing costs of travel to larger centres - the same kind of mechanism we saw in the Poston and Wilson example in Chapter 4. This behaviour is represented in a spatial interaction model in which the location and size of shopping centres is given exogenously. Such a model can then be used to predict for example, the total revenues attracted to each centre.