Thursday, April 22, 2010

The Delusion and Deception in Rail Forecasts

Professor Flybjerg of Oxford University has made a career in proving the fallacies involved in the planning and management of megaprojects, and in providing ways to avoid major pitfalls and deliver projects with transparency and accountability. Of course his advice is ignored by Honolulu city administration and its consultants who follow the traditional "Community Railroading Handbook" where smoke-and-mirror shows and media commercials provide grossly inflated views of project benefits and hardly mention any of the disadvantages.

Proponents of megaprojects are both deluded and deceptive. Delusion and deception are two fundamental reasons why things go wrong with large projects. Every forecast contains an amount of delusion and deception. It comes with the territory because of bias and future uncertainties.

We are all familiar with weather forecasts. Only a small amount of delusion and deception is involved there, as shown in the figure below. Meteorologists have no desire to deceive anyone about the weather; but sometimes they use words that brighten the picture or accentuate the risk. Meteorologists are not deluded because they consult with a comprehensive weather model; but the weather model has some built-in inaccuracies which lead to some wrong predictions.
In Flybjerg's analysis, the weather forecast is a best case scientific prediction when it comes to deception and delusion. What are the worst? Top worst are unique projects like the Sydney Opera House and our infamous all-steel Aloha Stadium. A close second worst is rail projects as seen below. Who would have thought that good ol’ steel wheel on steel rail projects would be so rife with delusion and deception. But data from dozens of actual projects show just that. The promoters of rail projects are deluded about ridership and cost forecasts. And the same promoters deceive the public about ridership and cost forecasts.

One thing we can say for Honolulu's administration and its consultants when it comes to the proposed rail is that they are consistent with past experience elsewhere. Their forecasts are full of delusion and deception. Here is some evidence.

Denver, a city with substantial rail construction expertise and a more balanced political climate ran a large deficit for its
FastTracks rail --planned to cost $6.6 billion, now it needs an extra $2.4 billion to get it completed as planned. On the other hand, Honolulu insists that its planned cost of $5.4 billion will hold. That’s without completing the environmental process and without the recent airport rerouting to clear the airport runway. Many cities two to five times larger than Honolulu cannot get more than 40,000 daily riders on their light rail systems after several years in operation. But Honolulu insists that on opening day its rail will carry 97,000 riders and some years later this will balloon to over 130,000. That's simply nuts.

How about
Tren Urbano of San Juan, Puerto Rico? This is a good comparison with Oahu because Tren Urbano is heavy rail, it is on an island, it is planned by the same consultant, it had the same Federal Transit Administration oversight, and it is only five years old.

The consultant forecast a cost of $1.25 billion. FTA approved it. The actual cost was $2.25 billion. Nearly double!

The consultant forecast 80,000 riders. FTA approved it. The actual ridership was under 25,000 and for months they had to run the trains for free. Now after five years it carries 38,700 daily riders. Not even half of the original forecast!

So what did the
Tren Urbano delusion and deception accomplish? It cheated the system itself. The project proponents cooked the numbers so that FTA's required cost/benefit value of no more than $25 per new transit rider is met. This is a critical estimate. If the cost projection is under $25, then a transit project is eligible for federal funding.

Let's say that the
Tren Urbano just made the FTA cut by having a cost ratio of $24.8. Once the system was built, all delusion and deception went away and San Juan suffered the results of a poorly planned project. The actual ratio came out to $92.3 or 270% higher than the approved ratio! With accurate forecasts, this system would never have a chance of receiving federal funding.

Tren Urbano, TheRail forecasts are full of delusion and deception. It's a good thing that “the train hasn’t left the station” and it ain’t leaving any time soon.