Abstract
In the paper the observation is made that in the economic literature one sided attention is given to either flow- or bottleneck congestion. Since both forms appear to occur simultaneously in practice, there seems to be room for an integral model. In the paper some steps in this direction are made. In doing this, use is made of the so- called fundamental graph which shows the relation between traffic density and volume. It is shown that hysteresis is a direct consequence of the usual curvature that is given to this graph.