Extract

To slightly misquote the late American political scientist, Ezra Solomon (1986, p. E1), ‘The only function of [traffic] forecasting is to make astrology look respectable’. While Solomon was, of course talking about economics, the sentiment extends to traffic and transportation forecasting in its many guises. Transportation forecasting, as we now understand it, is relatively new, but even an early assessment of traffic forecasting in Britain did not paint a very complementary picture of the engineering-based methods used at the time: ‘an assumption of zero change from the base year would not have produced larger forecast errors . . . for trips the errors would have been considerably less . . . [than those obtained from a transport model]’ (Mackinder and Evans, 1981, p. 25). We are simply not very good at it or, in some instances we allow vested interests to manipulate predictions to achieve their own self-serving ends. The practical challenges of producing good forecasts have perhaps grown over time. Most forms of modern transportation require a substantial amount of expensive, durable, and inflexible infrastructure to be efficient. This infrastructure embraces not only roads, ports, rail lines, and the like that is directly used for movement, but also specialized facilities to build airframes, cars, ships, and railway units. It is also a sector that has traditionally been the subject of significant regulation, generally for economic, military, and political reasons but more recently for environmental protection.