Measuring the accuracy of bus rapid transit forecasts

Abstract

The research of Dr. Bent Flyvbjerg in the 1990s and early 2000s showed that urban rail projects often cost more than estimated and carried fewer riders than projected, a troubling trend suggesting that the forecasts for urban rail projects were too optimistic in terms of cost and ridership. Inspired by that research, this analysis seeks to extend that framework to analyze Bus Rapid Transit (BRT). A study of forecast vs. actual costs and ridership was conducted for 19 BRT projects in the United States. From this, it was found that the cost projections for these projects tended to be quite accurate, but ridership projections tended to be quite inaccurate and showed a clear tendency towards an optimism bias. As BRT becomes a more common choice for rapid transit investment in the US, this analysis suggests that current ridership forecasting methods still leave much to be desired.