Can induced traffic be measured by surveys?

This paper addresses the feasibility of measuring induced traffic by means of appropriately designed surveys. The problems which confront any attempt to measure induced traffic are described and discussed. They include: inherent variability of traffic data, difficulty in establishing what would have happened in the absence of the scheme, uncertainties in the attribution of cause and difficulty in determining the most appropriate time to conduct surveys. Some of these problems have no solution while others can be reduced with an appropriate programme of surveys and control studies. Calculations are made of the sample sizes required to achieve various levels of target accuracy. It is concluded that, given an appropriate programme of traffic counts including control studies and extensive screenlines, it should be possible at relatively modest cost and with reasonable precision, to measure the increases in traffic associated with a scheme and to identify how much of that increase is due to rerouteing. Any further disaggregation of this extra traffic into that which is due to change of mode, change of destination and increased frequency, is much more problematic. The basic traffic counts would need to be supplemented by a programme of public transport surveys, registration plate marking and/or roadside interviews, which would add very considerably to the cost of the exercise. However, even with considerable expenditure, adequate precision may be difficult to achieve and attribution of cause impossible to make.