Rail passenger forecasting called into questionPublished 24 February 2016 by Jamie Wylie
We have this week issued a reply to Scottish Government cabinet secretary Keith Brown MSP following correspondence regarding rail passenger forecasts.
Whilst we welcome recent positive developments in rail, in particular the Borders Railway link, room for improvement remains in assessing passenger demand and usage in various routes across Scotland.
Commenting on Mr Brown’s response, Transform Scotland director Colin Howden said:
“While the report states “The comparison of forecast and observed demand showed that there is no systematic under or over-forecasting of demand for new stations.” (Section 3.70), it then goes on to indicate that “For a number of stations (Alloa, Edinburgh Park and Newcraighall) there was insufficient information available on how the forecasts had been prepared, and the reason for the under or over-forecast could only be surmised.” (Section 3.71). This is borne out by the findings reported for Larkhall, Edinburgh Park and Alloa where the outturn passenger numbers exceed those forecast by 21%, 83% and 180% respectively. The TS/DfT report is also now six years old, and hence reports no evidence on subsequent successful Scottish rail reopenings such as the Airdrie-Bathgate Line and Laurencekirk Station. Furthermore, initial experience from the Borders Railway would also support a pattern of substantially under-estimated demand for Scottish reopening schemes, with the actual demand for the stations in the Borders (Stow, Galashiels and Tweedbank) far above that forecasted.
“The TS/DfT report made the recommendation that demand forecasting could achieve better forecasting results by using the ‘Generic Station Forecasting Model’. We are increasingly of the opinion that new techniques may be required for rural / regional routes. It may be that recent Borders Railway experience renders even the ‘Generic Station Forecasting Model’ redundant. It is therefore crucial that the forthcoming official feasibility study for taking the Borders Railway onwards from Tweedbank to Hawick and Carlisle ditches the flawed methodology used to forecast patronage of the Borders Railway. Traditional modelling techniques for classic commuter routes may not be appropriate for longer rural and regional railways with strong leisure and tourism potential.”