The argument has been made that, with a few exceptions, travel demand management (TDM) measures have had little success and at most this has been local. An important assumption underpinning the application of TDM measures directed at changing vehicle occupancy, trip timing and route choice behaviour is that the traffic stream has a low level of variability from day to day because travel behaviour is habitual. In apparent contradiction to this assumption are the findings from traffic surveys that while traffic volumes are reasonably consistent, there is a high variability in those travelling on different days. This could explain why TDM measures have not been successful, either because the variability was too large for the change to be significant, or because the measures that were applied were based on incorrect assumptions.
The authors presented a paper at the 2006 SATC conference which reported the results of a licence plate recognition survey on an arterial route entering the Cape Town CBD. It reported levels of variability which, while consistent with similar international studies, were surprisingly high. This paper will describe the results of two follow up studies of traffic variability in two different environments; namely an arterial road in the suburbs, and a residential collector. For each study, vehicle registrations were noted for 90 minutes in the morning peak period on the week day of three weeks. These were compared to determine the variability within the traffic streams.
A comparison of the three datasets reveals that there is a significant difference in the degree of variability in the traffic streams of the three different environments. This significant difference provides an opportunity to reconsider the appropriateness and potential for TDM measures that assume habitual travel behaviour; more specifically those aimed at changing trip timing and route.