- 9 out of 10 programmes have cost overrun.
- Overruns of 50-100 percent are common.
- Overruns above 100 percent are not uncommon.
- Overrun is everywhere; no nation, sector or programme type is without overrun.
- Overrun is constant for the 70-year period for which comparable data exist; cost estimates have not improved over time.
- For documentation on cost overrun, click here and here.
- Only few people and organizations have a track record of being able to repeatedly manage major projects on budget and time. But such people and organizations do exist and are important to learn from. For more, click click here.
Examples of programmes with large cost overruns:
- Suez Canal, Egypt , 1,900 percent.
- Sydney Opera House, Australia, 1,400 percent.
- Concorde supersonic aeroplane, UK, France, 1,100 percent.
- Boston's Big Dig, USA, 275 percent.
- Britain's West Coast Main Line rail upgrade, 260 percent.
- Denver international airport, USA, 200 percent.
- Panama canal, 200 percent.
- Humber bridge, UK, 175 percent.
- Copenhagen metro, Denmark, 150 percent.
- San Juan Tren Urbano, Puerto Rico, 135 percent.
- Northeast Corridor rail line, USA, 130 percent.
- Great Belt rail tunnel, Denmark, 120 percent.
- Channel tunnel, UK, France, 100 percent.
- Shinkansen-Joetsu rail line, Japan, 100 percent.
For benefit shortfalls, valid and reliable data are even harder to find than for cost overruns. Only few studies exist with statistically valid conclusions. One such study, of passenger demand shortfalls on new rail lines, shows:
- 9 out of 10 rail programmes have passenger demand shortfalls.
- The average shortfall is 51.4 percent, equivalent to an average overestimate in rail passenger forecasts of no less than 105.6 percent.
- Only 16 percent of demand forecasts are within ±20 percent accuracy.
- No improvements in shortfalls and estimates have been detected for the 30-year period for which data exist.
- For documentation on benefit shortfalls, click here and here.
Examples of programmes with large demand shortfalls:
- Calcutta metro, India, 95 percent.
- Miami metro, Florida, USA, 85 percent.
- Channel tunnel, UK, France, 82 percent.
- Paris Nord TGV line, France, 75 percent.
Demand shortfall is here measured as the percentage of predicted passengers that never showed up on the trains during the first year of operations. The shortfall is measured against the prediction of the final business case.
For motorways and highways the results are more balanced, i.e., the number of roads with overestimated and underestimated travel demand, respectively, is about the same. Even so, half of road traffic forecasts are wrong by more than ±20 percent.
The following figure shows rail projects, for which data was available for both cost overrun and demand shortfall:
Each dot in the figure is a rail project. A project with 1.0 on both the x-axis and y-axis would be exactly on target with no cost overrun and no demand shortfall. A project with 1.5 on the x-axis and 0.5 on the y-axis would have 50 percent cost overrun and 50 percent demand shortfall. Cash cows, i.e., projects with lower costs and higher demand than projected would be in the upper left-hand corner of the figure, but there are none. Financial disasters – projects with higher costs and lower demand than projected – are in the lower right-hand corner, where we find the majority of projects. Even with a generous definition of being "on target" that accepts deviations of ±20 percent, as indicated by the dotted lines, we see that not a single project was on target for both cost and demand.
The combination of cost overrun and benefit shortfall is toxic to programme viability, but is nevertheless common, and not only for rail programmes (for more click here).
For more about cost overruns, click here:
Wikipedia, Google Scholar, Google, Bing.
For more about benefit shortfalls, click here:
Wikipedia, Google Scholar, Google, Bing.