“Federal, state, and local governments are banned from creating free roads to compete with toll roads.” That’s another claim being made by some conservative and libertarian opponents of toll roads. Some claim that such “non-compete” provisions were invented by private toll operators and are now being copied by public-sector toll roads.
What are the facts?
First, a bit of history. Most 20th century toll roads were developed and operated by government toll authorities. When they went to the bond market to sell toll revenue bonds for brand-new toll roads, bond underwriters asked them a logical question: “What happens to our annual debt service payments if 15 years into our 30-year bond your state highway department builds a shiny new ‘free’ highway parallel to our toll road, and half of your toll-payers switch to that?”
In many cases, the toll agency and the highway department conferred and reached an agreement that the highway department would not build a parallel highway during the life of the bonds. Such “non-compete” agreements were not always used, in part because with ever-present funding limitations, state highway departments had better things to do with their money than duplicating the occasional toll road.
Now, let’s fast forward to 1991 in California. The state has launched a pilot program—the first in the nation—to test the idea of public-private partnership (PPP) toll roads. The top-ranked proposal calls for the private consortium to finance, build, and operate express toll lanes in the median of one of the most congested freeways in the state: SR 91 in Orange County. The draft financing plan is presented to the bond houses and they react in horror. “You propose building a toll road literally right next to a freeway,” they say. “What if the government widens the freeway and takes away half your toll-payers?” The company took this news to Caltrans (the state DOT) which, at the time, had decided that adding the toll lanes was to be the last expansion of SR 91. So Caltrans figured it was not giving up anything by agreeing to a non-compete provision in the long-term contract with the consortium. That satisfied the bond houses, and the project was financed and built.
Although the toll lanes were a big success, growth in the region was faster than expected, so less than 10 years after the new lanes opened, the regular lanes on the freeway were congested again, and voters started agitating for more regular lanes. Because Caltrans could not add lanes, due to the non-compete provision, the outcome after several years of debate was that Orange County bought the toll lanes and eliminated the non-compete provision.
Modest capacity additions have been made since then, but the toll lanes are still operating in the black. In hindsight, we can see that the demand for congestion relief in that corridor was so strong that the non-compete provision was superfluous.
In the subsequent decade, nearly a dozen public-private partnership toll road projects have been financed around the country. Most have not included non-compete provisions. A few (such as the Chicago Skyway) have no provisions of any kind on this subject.
More common is what we might call second-generation competition provisions. Instead of preventing the state from building new (potentially competing) highways, these provisions allow for some degree of compensation to the toll company for lost toll revenue, if the state builds a road that competes directly with the toll road or lanes.
In 2007 Texas clarified its PPP law in this area. It prohibits true non-compete provisions like the one that once existed for SR 91. There can be no agreements that forbid a state agency to build any transportation project. But long-term PPP agreements may provide for some degree of compensation for lost toll revenue under limited conditions: if the state builds a new highway that is not in its long-range transportation plan but is within 4 miles of the PPP toll road. Similar provisions exist in other states.
In the world we live in, this strikes me as a reasonable compromise. PPP toll roads are going forward mostly in high-growth states with huge shortfalls in transportation funding. In those states (like California, Florida, Texas, and Virginia) gas tax revenues can barely cover the cost of maintaining existing highways, leaving very little money to invest in the additional capacity needed to cope with growth.
Using revenue bonds backed by toll revenues, the private sector can add needed lanes and highways, on a user-pays basis. But these are inherently risky investments, and part of that risk is political risk—that the government will change the rules after investors’ money is embedded in immovable concrete, steel, and asphalt. Toll road companies can take into account the projects government plans to build over the next 25 years; that’s why those projects are excluded from any compensation possibility. It’s only the unexpected competing road that some future politician might dream up to “get back at” the toll road company that would be subject to compensation if the toll company can demonstrate lost revenue.
Actually, the best protection against future predatory action by government is revenue-sharing provisions. Increasingly, long-term PPP toll road agreements include provisions for sharing a portion of toll road revenues with the relevant government, if the toll road does better than its base-case forecast. That kind of provision aligns the interests of both parties to the public-private partnership, over the full life of the agreement.