5.8.07

THE CANADIAN NATIONAL AND THE CANADIAN PACIFIC. Andrew Sullivan has some guest writers this week, and two of them have been looking at privatization of the roads as a way to ensure that owners will take care of the property. The idea is not without merit, provided one gets the incentives right. The current shabby state of our roads cannot be attributed to only one cause, but a combination of political unwillingness to charge vehicle operators prices that reflect the wear and tear and the congestion costs (yes, I'm looking at that 53 foot trailer over there, and that Escalade speeding in the fast lane) and political willingness to commingle fuel and tire tax funds with general revenues to make the federal deficit look smaller is destructive, and the propensity of elected representatives to appropriate money for visible but not necessarily useful projects is inefficient. Perhaps some market tests are in order.

One post links to an Ideoblog (appears to be based in Illinois) reflection on potential pitfalls with a private road operator.
Once the contractor gets the road, there's a risk it will be as bad a caretaker as the government. Malanga notes the "significant restrictions and operating requirements written into the contracts," like the one for the Indiana tollroad. But how do we know government agents will enforce these provisions?
That's the sticking point with Demsetz auctions (an alternative to rate-of-return regulation that is being used, in modified form, by Britain's Network Rail to obtain passenger train operators.) The highway, like an electrical grid or a railroad line, is a specific asset that can be wasted by the operator, absent sufficient safeguards for proper performance (and the British are still struggling with the proper safeguards when several franchise holders are operating trains on the same tracks) and the incumbent operator can learn things in the act of running the property that will confer it an advantage over other bidders when the franchise is rebid.

A second post takes a skeptical view of any such privatization.
Privatization of highways doesn't really solve that; it just replaces a public monopoly with a private monopoly. And since the whole point of private ownership is that superior products will result through competition, how is any problem actually solved here?
That's relatively easy to fix. The competition policy must provide incentives for multiple potential operators of any one road, as well as opportunities for the franchise to be rebid on a regular basis. There's no difference between selling the right to operate Interstate 40 and selling the right to operate Inter-City Cross Country.
And if competition is to be introduced, offering a similar road nearby, the barriers to entry are enormous. Competition would have to come through a rival company building a whole separate road nearby, first buying up and developing land that could have provided housing, businesses, green space, etc. It would take years to raise the business capital, then take care of all that overhead and construct the separate road, and finally open for business. That sort of situation would hardly put much competitive pressure on a current monopoly holder.
The point of the Demsetz auction is to introduce competition for the right to run the road. Presumably the bidders would be required to offer a schedule of tolls sufficient to cover the costs of maintaining the road, with the contract awarded to the lowest bidder, and with proper arrangements to indemnify the road commission in case the company has to quit the franchise.

On the other hand, if we were to experiment with road privatization, I know exactly where I'd try it. New Jersey has two separate toll roads that each run the length of the state from north to south, the Garden State Parkway and the New Jersey Turnpike. They're fairly close to each other, and would be redundant if not for the fact that the state's traffic needs are so severe.

It might not be a bad idea to quasi-privatize the two highways via indefinite leases, with the condition that the two roads can never come back under the same ownership — any attempt to do so would result in the forfeiture of ownership back to the state government. That way, two companies would have to maintain their roads and offer decent tolls in order to attract drivers to come to their highway as opposed to the other. And since these two separate roads already exist as it is, the barriers to entry don't apply.

That's not a bad idea. In fact, New Jersey could start by privatizing the Turnpike and continuing to operate the Parkway. Railroad deregulation in North America actually started in Canada (using the locution "commercial freedom") and both the (now-privatized) government owned railroad and the privately owned railroad obtained productivity gains after deregulation. The idea applies more generally. For example, why not have one operator for Interstate 70 and another operator for Interstate 80 across Indiana, Ohio, and Pennsylvania? And why not sell some franchises that include the possibility of upgrades, such that the franchise operator upgrades U.S. Highway 6 (east-west) or 45 (north-south) as limited-access turnpikes? There's a great deal more room for competition among highway operators than there is among railroads, where we currently have two systems coast-to-coast in Canada with substantial presence in the Lower 48, and two eastern and two western systems in the United States, with the western carriers having affiliates in Mexico.

No comments: