Construction cost and traffic performance of transport infrastructure projects: Evidence from a multilateral lender

Source abstract

: This paper presents an empirical analysis of forecast errors in transport infrastructure projects appraised and supported by the European Investment Bank (EIB), the world’s largest multilateral lender. Evidence of systematic bias, found in some earlier literature in other contexts, would have implications for the practice of project appraisal. While recognising the inherent uncertainty in forecasting, we do not find any systematic and significant evidence of bias in the forecasts made by EIB transport sector experts of construction costs, nor are there differences between transport sub-sectors in terms of the consistency (precision) of cost forecasts. There is some limited and mixed evidence of an upward bias in the forecasts of first-year traffic in some, but not all, transport sub-sectors, but there are no significant differences in precision across them. The distribution of forecast errors in both construction costs and first-year demand are skewed and have outliers. The outliers appear mainly related to factors internal to projects such as scope changes and to their external economic environment. There is no conclusive evidence that the Covid pandemic would have led to any bias in the forecasts or affected their precision, but there is some, if inconclusive, evidence that traffic forecasts in projects outside the EU are subject to systematic shortfalls and less precision than inside the EU. These results differ from some earlier literature, but they are qualitatively similar to more recent empirical analyses. We discuss the role of institutional factors and incentives in driving our results.