Traffic and Revenue Forecasting Overestimates: The Case of Colorado's Northwest Corridor

Less than one year after the completion of the "final" transportation and environmental study for the Northwest Corridor near Golden, CO, consultants for the community from CRA International wrote a report noting that, in the end, only 6% of the total $922MM construction cost of the road could be financed.  The following are excerpts from the executive summary.  The full report can be read at "Report and comments on the Northwest Corridor Transportation and Environmental Planning Study. Published July 2008."

And, by the way, Wilbur Smith Associates prepared most of the traffic and revenue forecasts used in the Northwest Corridor Transportation and Environmental Planning Study, July 2008, that is critiqued by the CRA International consultants. 

We conclude that the TEPS’s (Transportation and Environmental Planning Study) financial analysis includes fundamental errors that cause it to overstate the amount of potential revenue and bonding capacity from the proposed “Combined Alternative” by a factor of at least 2-3 times. These errors render the TEPS financial and traffic analyses useless without correction, as discussed below. The result of completely correcting all of the errors other than a large underestimate of likely operations and maintenance (O&M) costs is that only 6% of the $922 million construction cost of the Combined Alternative could be financed from the toll revenue bond proceeds. And if the toll road O&M cost is based on even only a portion of the Northwest Parkway’s most recent publicly available operating and maintenance cost, none of the construction cost can be financed with revenue bonds. . .

This analysis identifies and quantifies the effects of six important problems with the analysis and results for the entire Combined Alternative financial analysis presented in the TEPS Report. Completely correcting all but one of the six fundamental problems reduces the TEPS study estimate of $135- $230 million construction bond proceeds by 57%-75% to $57.7 million, or only 6% of the $922 million construction cost of the Combined Alternative. Completely correcting all the problems means that none of the construction cost can be financed.

We identified the magnitude of these problems by rerunning the DRCOG model and duplicating the TEPS study revenue bond calculations and changing the input values to account for the effects of each problem. The six problems, with their individual effects on the construction bonding capacity are:

1. The TEPS study failed to subtract O&M costs from toll road revenues in all estimates but the low-range TEPS study estimate. This renders estimates higher than the $135 million low range unrealistic and unusable.
2. The TEPS report drastically underestimated future operating and maintenance expenses for the toll road for even this “low range” estimate by holding them essentially constant in real dollars, even though the report acknowledges that they should increase significantly in the future. Partially correcting this error leads to a 23% reduction in potential bond proceeds. And, if the TEPS study used the most recent publicly-available O&M costs for the Northwest Parkway, there would be no revenue bond proceeds.
3. The TEPS report overestimated the number of highway trips by using freeway trip tables rather than the appropriate trip tables for the toll road and arterials making up the Combined Alternative. Correcting this error by itself leads to a 37% reduction in potential bond proceeds.
4. The TEPS report’s modeling also overstates traffic and revenue by coding the portion of the Combined Alternative on US 6 in Golden as a six-lane freeway rather than a regional arterial. Correcting this error results in an 11% reduction in potential bond proceeds.
5. The TEPS report also relies on borrowing costs from almost three years ago that are 0.5 percentage points lower than current costs. Correcting these costs to current borrowing costs reduces the potential bond proceeds by 10%.
6. The TEPS financial analysis also unrealistically assumed an immediate initiation of full traffic on the toll road without a ramp-up period, contrary to the experience of every new toll road. Inclusion of even a very optimistic ramp-up period would reduce potential bond proceeds by 6%.

The conclusion of this analysis is that the TEPS study grossly overestimates the traffic, revenue and construction cost bonding capacity for the Combined Alternative, including the section of toll road now called the “Jefferson Parkway.” Given the problems we describe in this analysis, even the chances of financing 6% of the $922 million construction cost of the Combined Alternative are remote, and not likely to be embraced by any investor. The 6% is well within the range of the statistical noise of the construction cost estimate.

Under these circumstances, there is no business case for attracting an investor to incur both the revenue risk and the construction cost risk. Further, the financial estimates of the TEPS report can not be used for financial planning or feasibility assessments without correction of its fundamental errors.

So what confidence should we have that Wilbur Smith has done a better job of forecasting future traffic and revenue on the Dulles Toll Road?  A second, independent traffic and revenue forecast is clearly required before making a multi-billion dollar commitment to continue construction on the Silver Line.