Privatizing the Nation's Highways--and Other Infrastructure--Posner
The State of Indiana has just leased the Indiana Toll Road--a 157-mile-long highway in northern Indiana that connects Illinois to Ohio--to a Spanish-Australian consortium for 75 years for $3.8 billion, to be paid in a lump sum. (The deal has been challenged in the Indiana state courts.) The lease is complex, imposing many duties on the lessee (such as to install electronic toll collection, in which Indiana has lagged). A key provision is that the consortium will not be able to raise toll rates until 2016 (for passenger cars--2010 for trucks) and then only by the greatest of 2 percent a year, the consumer inflation rate (CPI), or the annual increase in GDP. (On the eve of the lease, Indiana raised toll rates--which hadn't changed since 1985--significantly.) Two years ago Chicago made a similar lease of the Chicago Skyway, an 8-mile stretch that connects Chicago to the Indiana Toll Road, for $1.8 billion. There is considerable interest in other states as well in leasing toll roads to private entities.
The idea of privatizing toll roads is an attractive one from an economic standpoint. Private companies are more efficient than public ones, at least in the limited sense of economizing on costs. I call this sense of efficiency "limited" because there are other dimensions of efficiency, for example the allocative; a monopolist might be very effective in limiting his costs, but by charging a monopoly price he would distort the allocation of resources. Some of his customers would be induced by the high price to switch to substitutes that cost more to make than the monopolist‚Äôs product but that, being priced at the competitive rather than the monopoly price, seemed cheaper to consumers. (This is the standard economic objection to monopoly.) The reason for the superior ability of private companies to control costs is that they have both a strong financial incentive and also competitive pressure to do so--factors that operate weakly or not all in the case of public agencies--and that their pricing and purchasing decisions, including decisions regarding wages and labor relations, are not distorted by political pressures and corruption. There is a long history of price-fixing in highway construction and maintenance, attributable in part to bidding rules that, in endeavoring to prevent corruption, facilitate bid rigging. For example, if to prevent corruption contracts are always awarded to the low bidder, a bid-rigging conspiracy will always know whether one of its members is cheating, if the low bidder, who gets the contract, was not the bidder that the conspiracy assigned to make the low bid. If cheating on a conspiracy is readily detectable, cheating is less likely and therefore the conspiracy more effective.
The problem of allocative efficiency looms when, for example, there are exernalities; but the solution to the problem rarely requires public ownership. One significant externality associated with vehicular transportation is the congestion externality: no driver is likely to consider the effect of his driving on the convenience of other drivers, because there is no way in which he can exact compensation from drivers for not driving or driving less and therefore improving their driving time. That externality is internalized by a toll road, because congestion reduces the quality of the driving experience and so the amount each driver is willing to pay in tolls; the owner of the toll road will trade that willingess to pay off against the reduction in the number of drivers as a result of a higher toll.
Another externality, however, will not be internalized by the toll-road operators. That is the contribution that driving makes to pollution and global warming. But public ownership is not necessary in order to internalize this externality. The government can force its internalizing by imposing a tax on driving.
There is, however, in the toll-road setting another source of allocative inefficiency, and that is monopoly, which I have mentioned already. Drivers who do not have good alternatives to using the Indiana Toll Road can be made to pay tolls that exceed wear and tear, congestion effects, social costs of pollution, and other costs of the road, engendering inefficient substitutions by drivers unwilling to pay those tolls. To an extent, the toll-road operator may be able to discourage substitution by price discrimination, but this is unlikely to be fully effective and indeed can actually increase the allocative inefficiency of the monopoly.
The monopoly issue raises the question: what exactly was Indiana selling when it leased the toll road for $3.8 billion? The higher the tolls and the greater the lessee's freedom to raise the tolls in the future, the higher the price that the state can command for the lease. If the lease placed no limitations on tolls, the state would be selling an unregulated monopoly. If the lease could constrain the lessee to charge tolls just equal to the cost of operating the toll road (including maintenance, repairs, snow removal, lighting, and the collection of the tolls), the market price of the lease would be significantly lower. To the extent that the state wants to maximize its take from the lease, it will be creating allocative inefficiency by conferring monopoly power on the lessee.
It is difficult to determine whether the $3.8 billion price tag for the Indiana Toll Road is closer to the competitive or the monopoly price level. On the one hand, the lessee cannot raise tolls until 2010 or 2016 (depending on the type of vehicle), and increases after that are capped. On the other hand, the tolls were raised significantly just before the lease, and allowing the operator in 2010 to begin raising toll rates annually by the increase in GDP may confer windfall gains, since the cost of operating the toll road may not increase at so great a rate. One would have to know a great deal more about the economics of operating a highway than I do to figure out whether the terms of the lease confer monopoly power on the lessee.
I do not regard the monopoly concern as a strong objection to the leasing of the toll road, however. The reason is that most, maybe all, taxes have monopoly-like effects, in the sense of driving a wedge between cost and price. Suppose the lease price would have been only $2 billion had the state imposed more stringent limitations on toll increases. Then the state would have $1.8 billion less in revenue and would presumably make up the difference by increasing tax rates or imposing additional taxes, and these measures would have allocative effects similar to those of higher tolls charged by the lessee of the toll road. If the monopoly issue is therefore considered a wash, the principal effect of the lease will be the positive one of reducing the quality-adjusted cost of operating the toll road and the lease is clearly a good idea.
Toll roads are more attractive candidates for privatization than non-toll roads because it is easy to charge user fees; tolls are user fees. It would be harder to charge for the use of city streets, though no longer impossible, given electronic technology for monitoring drivers. Privatizing certain security services pose special problems as well, as Becker and I discussed in our May 28 posts about security contractors in Iraq. But public services the cost of which is defrayed in whole or significant part by user fees are good candidates for privatization, including Amtrak, the Postal Service, building and restaurant inspections, veterans' hospitals, and federal, state, and local airports. The privatization movement has a long way to go before achieving an optimal mixture of public and private service providers.
Against all this it will be argued--it is an argument emphasized by opponents of leasing the Indiana Toll Road--that privatization, at least when it takes the form of a sale or long-term lease of government property for a lump sum, beggars the future by depriving government of an income-producing asset. The argument, at least in its simplest form, is unsound, because the state is not disposing of an asset but merely changing its form: from a highway to cash. The subtler form of the argument is that, given the truncated horizons of elected officials, the state will not invest the cash wisely for the long term, but will squander it on short-term projects. This is a danger--how great a one I do not know. It would be an interesting study to trace the uses to which privatizing governments here and abroad have put the proceeds of sales of public assets.