The term "infrastructure" is much used but lacks a clear definition. I shall use it to mean inputs, often provided by government rather than private enterprise, into a very large variety of products and services. Good examples are transportation, communications (including the Internet), education, the environment (including water resources), public health (in the sense of prevention of communicable diseases), and law enforcement (including the judiciary). Current concerns with our allegedly deteriorating infrastructure focus on road and air transportation and on primary and secondary education, and I shall confine my discussion to them.
I am going to bracket road transportation and education because the problems of both could probably be solved satisfactorily by privatization or, in the case of education, semi-privatization. Air transportation I discuss separately because its problems almost certainly require a governmental solution.
America's roads (including bridges, which carry mainly roads) have been deteriorating, as is obvious to any user of the interstate highway system. The reason for the deterioration is that the system is carrying vastly more traffic than it was designed for in the 1950s. The result is not only rough surfaces which wear out tires and slow down traffic, but also delays due both to construction and to the sheer increase in traffic volume. Congestion and therefore potholes and delay have also increased on local and commuter roads.
Wear and tear, and delay, are real costs, but the costs of road building and road improving to reduce the costs of wear and tear and of delay are real too, and the challenge is to spend only up to the point where the last dollar spent yields a dollar in benefit. Government seems incapable of doing that. Privatization should be able to. This is obvious in the case of toll roads, where in fact a privatization movement is under way, as we have discussed in a prior post. The toll can be set equal to the cost that a vehicle imposes on the road in congestion and wear and tear. (That cost varies with the size and weight of the vehicle, but the toll can be made to vary with these factors as well.)
There is a concern about monopoly; there are not always good alternatives to a particular route. If the state or other public owner of the toll road auctions the road to the highest bidder, the winning bid will capitalize monopoly rents, and the tolls will therefore contain a monopoly markup. That is inefficient, but the revenue that the state obtains from the auction is a tax substitute, and taxes have the same general misallocative effects as monopoly prices. The state can if it wish avoid the monopoly problem by specifying a minimum quality of service and then auctioning the road to the lowest bidder who agrees to provide that quality.
Not all roads are toll roads, but given modern technology all can be made toll roads. Electronic toll systems are available that do not require vehicles to slow down, and these can be adapted even to local roads, so that the entire street system of a city or an entire metropolitan area could be privatized. This may seem a bizarre suggestion, but it is no more bizarre that allowing a private company to own the telephone or cable grid of a city. The street system is just another grid, and the potential monopoly problems can be dealt with in the same way that cities deal with telephone or cable monopolies, where auctioning franchises may again be the most efficient approach.
The American system of public education is heavily criticized. Costs per pupil are high relative both to private education and to public education in foreign countries that turn out better students. Many parents are voting with their feet, as it were, by putting their kids in private (including parochial) schools, home schooling the kids, or moving to communities that have better public schools. These are alternatives are costly to parents. It is not clear why government is in the business of operating any educational facilities. It is appropriate for government to require that children attend school up to a specified age, to fix minimum educational standards with regard both to curriculum and to performance, and to finance the costs of education for impecunious parents. None of these things requires that government own and operate schools. But public education is not about to be abandoned, and it could be semi-privatized by adoption of a voucher system, which by permitting parents to choose among public schools would force public schools to compete with each other.
Air transportation presents a baffling problem in infrastructure deterioration. The deterioration in airline service in the last five years has been dramatic, involving as it does not only extraordinary delays but also horribly crowded airplanes and crummy airports. The delays are masked by the fact that the airlines have increased the scheduled time of flights and by the difficulty of measuring delay resulting from canceled flights and missed connecting flights. A study this past month by the majority staff of the Joint Economic Committee of Congress entitled "Your Flight Has Been Delayed Again" estimates the annual cost of airline delay at some $41 billion. Of this amount $12 billion is attributed to traveler time costs. That estimate strikes me as too low. It includes the delay due to schedule change, but excludes delays due to missing connecting flights, to canceled flights, and to lost time when in order to avoid missing an appointment one has to take an earlier flight and hence if it is not delayed has dead time on arrival. The uncertainties of air travel also cause some potential travelers to substitute, at some cost, another activity (maybe even another job, requiring less travel). On the other hand, the $41 billion figure is misleading because it is the cost of total air transportation delay, not avoidable delay; optimal delay is not zero, because of unavoidable weather conditions and equipment failures.
The deterioration in airline service is puzzling because the high cost of aviation fuel has virtually bankrupted much of the industry, and so one would expect a contraction; this is beginning yet there is no expectation that flight delays will diminish. The reason is that the airline industry has very heavy fixed costs, so that even when high fuel prices push up its marginal costs, each flight, provided the revenue from the flight exceeds the marginal cost of the flight, will contribute something to the airline's fixed costs, and so airlines are reluctant to reduce the number of their flights. This explains, moreover, why airline service has deteriorated. When demand grows, as it has done rapidly in recent years, the industry responds by adding flights. No airline has an incentive to balance the revenue from additional flights against the costs in additional delay, because one airline's reducing the number of its flights would have little effect on delay, but a dramatic negative effect on its revenues.
One culprit in the deterioration of airline service is Congress, and another is the Administration; between them, they have failed to create a modern air traffic control system that would reduce delay by reducing the safety-required spacing of planes both on the runway and in the air. But that would not solve the basic problem, which as I have said arises from the fact that no individual airline bears the full costs of the delays it creates. The airways are like a highway of fixed size with no tolls, facing an increase in traffic.
Another apparent culprit is the airports, but they are in much the same position as the airlines. They are locally owned and in principle (but not in practice) can control congestion by limiting the number of takeoffs and landings either directly or by fees. If every city had two airports, if most airline traffic were between just two cities, and if two companies each owned one airport in each city, then the competitive situation would be identical to that of two competing toll roads, and, assuming a modern air traffic control system, the optimal amount of delay would be achieved. But these conditions are not satisfied. No single airport can optimize congestion, because it does not control the traffic to and from, and hence delay at, other airports, and that delay will in turn cause delay at its airport.
The best solution might be a federal airplane congestion tax that would vary from route to route depending on delays, which vary considerable across regions and specific airports. The revenues could be used to update the air traffic control system. An alternative might be to allow some limited collusion among the airlines, enabling each airline to reduce the number of flights without losing business to its competitors. But that would be in essence a return to the system of airline regulation prior to the abolition of the Civil Aeronautics Board, and that was a thoroughly unsatisfactory system.