Bias in traffic forecasts lead to failing toll roads
Optimism Bias and Risk in Public Private Partnerships
The tolling technology is better than ever — but traffic forecasts are a disaster.
by James A. Bacon
Bacon’s Rebellion
November 14, 2014
Randy Salzman, a free-lance Charlottesville writer, has spent the last couple of years trying to understand how Public Private Partnerships (P3s) work in Virginia. If the private sector is supposed to be so much more efficient than government, he asks, how come so many big P3 transportation projects in Virginia and across the nation have gone bankrupt? Why do private sector companies continue investing in similar projects despite the obvious risk? And what exposure do taxpayers when deals go bad? He doesn’t have any definitive answers, but he lays out a lot of good questions in the latest issue of Style Weekly.
Salz, an occasional contributor to Bacon’s Rebellion, gets closest to the truth when he mentions the “optimism bias” in traffic forecasts. In project after project across the country, private P3 companies and their government partners have over-estimated traffic volumes on the roads they build. Writes Salz:
One study found that the projections tended to be 109 percent more than actual traffic — or more than double — and that nowhere in completed American P3s have actual traffic and toll income come close to projections.
Here in Virginia, flawed traffic forecasts were at the root of the Pocahontas Parkway debacle in eastern Henrico County and, if I’m not mistaken, the Dulles Greenway bankruptcy in Loudoun County (although that was not a P3 project). And there’s a very good chance that the Capital Beltway Express’s Northern Virginia HOT lanes project will experience a similar fate.
I think there are two things going on here. First, the private sector’s flawed traffic project models paralleled flawed public sector models. Everybody in the transportation business extrapolated the growth trends of the ’60s, 70s, ’80s and ’90s indefinitely into the future. I warned a decade ago that that was folly, but not many people listened. Reality set in in the mid-2000s when growth rates started tapering off and during the 2007-2008 recession, when traffic volume actually declined. The reasons are many and complex, as I have enumerated ad nauseum on this blog, but they are fundamental and lasting, not just a blip. We will not in our lifetimes return to the traffic-volume growth rates experienced during the post-World War II era.
The forecasts of traffic volume and associated toll revenues for the P3 projects were predicated on the assumption, now revealed to have been astonishingly naive, that traffic volume would increase on the same trajectory pretty much forever. That’s why the bankruptcies ensued, and why there will be more to come.
If experience tells us anything, the private sector will figure that out before the public sector does. As Salz quotes Lane Construction as saying in regard to proposed Interstate 66 toll lanes near Washington: Traffic projections have an “optimism bias.” Which brings us to the second reason for the wave of bad deals. Once someone, whether a private investor or a government agency, invests hundreds or thousands of man hours in analyzing a project, they get personally invested. No one likes to pull the plug. They want to see the project move forward. They tend to adopt assumptions that will make the project look more viable in order to obtain the financing needed to move it from paper to reality. This bias is so endemic in all types of projects that we can almost call it a part of human nature.
The private sector has built-in bullshit detectors. They’re called investors and bond holders. Investors want to generate a positive risk-adjusted return on investment. Bond-holders want to get their money back, plus interest. They may rely upon flawed traffic projects that no one questions, but they don’t suffer from the optimism bias of the project sponsors. They are naturally skeptical and have an interest in asking tough questions. Now, these investors and bond holders aren’t infallible by any means. They make bad investments, too. But they demand a higher standard of certainty than, say, politicians who want the glory of building a road but won’t be around to take the blame if the project falls apart.
Every toll-backed P3 project sells bonds to investors. How, then, did so many go wrong? The key is to look at how the public partner biased the outcome through subsidies and loan guarantees. Every big P3 project applies for financing from the federal Transportation Infrastructure Finance and Innovation Act (TIFIA). These federally guaranteed loans create a tranche of subordinated debt that creates a layer of protection for private bond holders. In other words, if Project A experiences a revenue shortfall, what revenues it does produce will go to bond holders first. Here’s how the Federal Highway Administration describes it: “The TIFIA lien on project revenues may be subordinated to those of senior lenders except in the case of bankruptcy, insolvency, or liquidation of the obligor.”
This layer of protection significantly reduces the risk for senior bond holders, who then demand fewer assurances than they would otherwise before purchasing the bonds. In Virginia, the commonwealth has reduced project risk by making significant cash contributions as well. Most of the P3 projects set up in Virginia in recent years have used some combination of TIFIA funding and public subsidies to make the projects work. Without these contributions, the perceived risk would have been far higher, and the chances of getting pure private financing would have been much diminished. It’s fair to say that many, if not most, of the deals never would have happened.
Combine these three factors — highly flawed long-term traffic projections embraced by the public and private sectors both, the optimism bias for specific projects, and the diminution of risk through TIFIA financing and public subsidies — and we can explain a lot of went wrong. That’s not an exhaustive list of explanations but it accounts for a lot.
There are some things that public-private partnerships do well. Capital Beltway Express, a partnership of Transurban and Fluor, did an exceptionally competent job on the design and construction of the U.S. 495 HOT lanes. A well-crafted P3 deal can offload the risk of project overruns and construction delays to the private-sector partner. P3s also encourage private players to bring innovative ideas to the table that the commonwealth may not have considered before.
But the private sector doesn’t have any free-market fairy dust that magically transforms an ill-conceived project into a money maker. A poorly crafted P3 can end up with state government getting clobbered if traffic projections and toll revenues fall short. A poorly crafted P3 can create something like the U.S. 460 fiasco in which the state stands to lose $300 million or more without a yard of roadway being built.
Salz is right to dig into these transportation mega-projects that can run into the billions of dollars. Virginia has a lot at stake, and neither he nor I are persuaded that the public or our elected politicians fully understand the issues. I hope he keeps digging.