DRI-211 for week of 3-29-15: Which First – Self-Driving Cars or Self-Flying Planes?

An Access Advertising EconBrief: 

Which First – Self-Driving Cars or Self-Flying Planes?

As details of the grisly demise of Lufthansa’s Germanwings flight 9525 gradually emerged, the truth became inescapable. The airliner had descended 10,000 feet in a quick but controlled manner, not the dead drop or death spiral of a disabled plane. No distress calls were sent. It became clear that the airplane had been deliberately steered into a mountainside. The recovery of the plane’s flight data recorder – the “black box” – provided the anticlimactic evidence of a mass murder wrapped around an apparent suicide: the sound of a chair scraping the floor as the flight crew’s captain excused himself from the cabin, followed by the sound of the cabin door closing, followed by the steady breathing of the co-pilot until the captain’s return. The sounds of the captain’s knocks and increasingly frantic demands to be readmitted to the cabin were finally accompanied by the last-minute screams and shrieks of the passengers as they saw the French Alps looming up before them.

The steady breathing inside the cabin showed that the copilot remained awake until the crash.

As we would expect, the reaction of the airline, Lufthansa, and government officials is now one of shock and disbelief. Brice Robin, Marseille public prosecutor, was asked if the copilot, Andreas Lubitz, had – to all intents and purposes – committed suicide. “I haven’t used the word suicide,” Robin demurred, while acknowledging the validity of the question. Carsten Spohr, Lufthansa’s CEO and himself a former pilot, begged to differ: “If a person takes 149 other people to their deaths with him, there is another word than suicide.” The obvious implication was that the other people were innocent bystanders, making this an act of mass murder that dwarfed the significance of the suicide.

This particular mass murder caught the news media off guard. We are inured to the customary form of mass murder, committed by a lone killer with handgun or rifle. He is using murder and the occasion of his death to attain the sense of personal empowerment he never realized in life. The news media reacts in stylized fashion with pious moralizing and calls for more and stronger laws against whatever weapon the killer happened to be using.

In the case of the airline industry, the last spasm of government regulation is still fresh in all our minds. It followed in response to the mass murder of 3,000 people on September 11, 2001 when terrorists hijacked commercial airliners and crashed them into the World Trade Center and the Pentagon. Regulation has marred airline travel with the pain of searches, scans, delays and tedium. Beyond that, the cabins of airliners have been hardened to make them impenetrable from the outside – in order to provide absolute security against another deliberately managed crash by madmen.

Oops. What about the madmen within?

But, after a few days of stunned disbelief, the chorus found its voice again. That voice sounded like Strother Martin’s in the movie Cool Hand Luke. What we have here is a failure to regulate. We’ll simply have to find a way to regulate the mental health of pilots. Obviously, the private sector is failing in its clear duty to protect the public, so government will have to step in.

Now if it were really possible for government to regulate mental health, wouldn’t the first priority be to regulate the mental health of politicians? Followed closely by bureaucrats? The likely annual deaths attributable to government run to six figures, far beyond any mayhem suicidal airline pilots might cause. Asking government to regulate the mental health of others is a little like giving the job to the inmates of a psychiatric hospital – perhaps on the theory that only somebody with mental illness can recognize and treat it in others.

Is this all we can muster in the face of this bizarre tragedy? No, tragedy sometimes gives us license to say things that wouldn’t resonate at other times. Now is the time to reorganize our system of air-traffic control, making it not only safer but better, faster and cheaper as well.

The Risk of Airline Travel Today: The State of the Art

Wall Street Journal Holman Jenkins goes straight to the heart of the matter in his recent column (03/29-29/2015, “Germanwings 9525 and the Future of Flight Safety”). The apparent mass-murder-by-pilot “highlights one way the technology has failed to advance as it should have.” Even though the commercial airline cockpit is “the most automated workplace in the world,” the sad fact is that “we are further along in planning for the autonomous car than for the autonomous airliner.”

How has the self-flying plane become not merely a theoretical possibility but a practical imperative? What stands in the way of its realization?

The answer to the first question lies in comparing the antiquated status quo in airline traffic control with the potential inherent in a system updated to current technological standards. The second answer lies in the recognition of the incentives posed by political economy.

Today’s “Horse and Buggy” System of Air-Traffic Control

For almost a century, air-traffic control throughout the world has operated under a “corridor system.” This has been accurately compared to the system of roads and lanes that governs vehicle transport on land, the obvious difference being that it incorporates additional vertical dimensions not present in the latter. Planes file flight plans that notify air-traffic controllers of their origin and ultimate destination. The planes are required to travel within specified flight corridors that are analogous to the lanes of a roadway. Controllers enforce distance limits between each plane, analogous to the “car-lengths” distance between the cars on roadways. Controllers regulate the order and sequence of takeoffs and landings at airports to prevent collisions.

Unfortunately, the corridor system is pockmarked with gross inefficiencies. Rather than being organized purely by function, it is instead governed primarily by political jurisdiction. This is jarringly evident in Europe, home to many countries in close physical proximity. An airline flight from one end of Europe to another may pass through dozens of different political jurisdictions, each time undergoing a “handoff” of radio contact for air-traffic control between plane and ground control.

In the U.S., centralized administration by the Federal Aviation Administration (FAA) surmounts some of this difficulty, but the antiquated reliance on radar for geographic positioning still demands that commercial aircraft report their positions periodically for handoff to a new air-traffic control boss. And the air corridors in the U.S. are little changed from the dawn of air-mail delivery in the 1920s and 30s, when hillside beacons provided vital navigational aids to pilots. Instead of regular, geometric air corridors, we have irregular, zigzag patterns that cause built-in delays in travel and waste of fuel. Meanwhile, the slightest glitch in weather or airport procedure can stack up planes on the ground or in the air and lead to rolling delays and mounting frustration among passengers.

Why Didn’t Airline Deregulation Solve or Ameliorate These Problems? 

Throughout the 20th century, the demand for airline travel grew like Topsy. But the system of air-traffic control remained antiquated. The only way that system could adjust to increased demand was by building more airports and hiring more air-traffic controllers. Building airports was complicated because major airports were constructed with public funds, not private investment. The rights-of-way, land acquisition costs, and advantages of sovereign immunity all militated against privatization. When air-traffic controllers became unionized, this guaranteed that the union would strive to restrict union membership in order to raise wages. This, too, made it difficult to cope with increases in passenger demand.

The deregulation of commercial airline entry and pricing that began in 1978 was an enormous boon to consumers. It ushered in a boom in airline travel. Paradoxically, this worsened the quality of the product consumers were offered because the federal government retained control over airline safety. This guaranteed that airport capacity and air-safety technology would not increase pari passu with consumer demand for airline travel. As Holman Jenkins puts it, the U.S. air-traffic-control system is “a government-run monopoly, astonishingly slow to upgrade its technology.” He cites the view of the leading expert on government regulation of transportation, Robert Poole of the Reason Foundation, that the system operates “as if Congress is its main customer.”

Private, profit-maximizing airlines have every incentive to insure the safe operation of their planes and the timely provision of service. Product quality is just as important to consumers as the price paid for service; indeed, it may well be more important. History shows that airline crashes have highly adverse effects on the business of the companies affected. At the margin, an airline that offers a lower price for a given flight or provides safer transportation to its customers or gives its customers less aggravation during their trip rates to make more money through its actions.

In contrast, government regulators have no occupational incentive to improve airline safety. To be sure, they have an incentive to regulate – hire staff, pass rules, impose directives and generally look as busy as possible in their everyday operations. When a crash occurs, they have a strong incentive to assume a grave demeanor, rush investigators to the scene, issue daily updates on results of investigations and eventually issue reports. These activities are the kinds of things that increase regulatory staffs and budgets, which in turn increase salaries of bureaucrats. They serve the public-relations interests of Congress, which controls regulatory budgets. But government regulators have no marginal incentive whatsoever to reduce the incidence of crashes or flight delays or passenger inconvenience – their bureaucratic compensation is not increased by improved productivity in these areas despite the fact that THIS IS REALLY WHAT WE WANT GOVERNMENT TO DO.

Thus, government regulators really have no incentive to modernize the air-traffic control system. And guess what? They haven’t done it; nor have they modernized the operation of airports. Indeed, the current system meets the needs of government well. It guarantees that accidents will continue to happen – this will continue to require investigation by government, thus providing a rationale for the FAA’s accident-investigation apparatus. Consumers will continue to complain about delays and airline misbehavior – this will require a government bureau to handle complaints and pretend to rectify mistakes made by airlines. And results of accident investigations will continue to show that something went wrong – after all, that is the definition of an accident, isn’t it? Well, the FAA’s job is to pretend to put that something right, whatever it might be.

The FAA and the Federal Transportation Safety Board (FTSB) are delighted with the status quo – it justifies their current existence. The last thing they want is a transition to a new, more efficient system that would eliminate accidents, errors and mistakes. That would weaken the rationale for big government. It would threaten the rationale for their jobs and their salaries.

Is there such a system on the horizon? Yes, there is.

Free Flight and the Future of Fully Automatic Airline Travel

A 09/06/2014 article in The Economist (“Free Flight”) is subtitled “As more aircraft take to the sky, new technology will allow pilots to pick their own routes but still avoid each other.” The article describes the activities of a Spanish technology company, Indra, involved in training a new breed of air-traffic controllers. The controllers do not shepherd planes to their destinations like leashed animals. Instead, they merely supervise autonomous pilots to make sure that their decisions harmonize with each other. The controllers are analogous to the auctioneers in the general equilibrium models of pricing developed by the 19th century economist Vilfredo Pareto.

The basic concept of free flight is that the pilot submits a flight plan allowing him or her to fly directly from origin to destination, without having to queue up in a travel corridor behind other planes and travel the comparatively indirect route dictated by the air-traffic control system. This allows closer spacing of planes in the air. Upon arrival, it also allows “continuous descent” rather than the more circuitous approach method that is now standard. This saves both time and fuel. For the European system, the average time saved has been estimated at ten minutes per flight. For the U. S., this would undoubtedly be greater. Translated into fuel, this would be a huge saving. For those concerned about the carbon dioxide emissions of airliners, this would be a boon.

The obvious question is: How are collisions to be avoided under the system of free flight? Technology provides the answer. Flight plans are submitted no less than 25 minutes in advance. Today’s high-speed computing power allows reconciliation of conflicts and any necessary adjustments in flight-paths to be made prior to takeoff. “Pilots” need only stick to their flight plan.

Streamlining of flight paths is only the beginning of the benefits of free flight. Technology now exists to replace the current system of radar and radio positioning of flights with satellite navigation. This would enable the exact positioning of a flight by controllers at a given moment. The European air-traffic control system is set to transition to satellite navigation by 2017; the U.S. system by 2020.

The upshot of all these advances is that the travel delays that currently have the public up in arms would be gone under the free flight system. It is estimated that the average error in flight arrivals would be no more than one minute.

Why must we wait another five years to reap the gains from a technology so manifestly beneficial? Older readers may recall the series of commercials in which Orson Welles promoted a wine with the slogan “We sell no wine before its time.” The motto of government regulation should be “we save no life before its time.”

The combination of free flight and satellite navigation is incredibly potent. As Jenkins notes, “the networking technology required to make [free flight] work [lends] itself naturally and almost inevitably to computerized aircraft controllable from the ground.” In other words, the human piloting of commercial aircraft has become obsolete – and has been so for years. The only thing standing between us and self-flying airliners has been the open opposition of commercial pilots and their union and the tacit opposition of the regulatory bureaucracy.

Virtually all airline crashes that occur now are the result of human error – or human deliberation. The publication Aviation Safety Network listed 8 crashes since 1994 that are believed to have been deliberately caused by the pilot. The fatalities involved were (in ascending order) 1, 1, 4, 12, 33, 44, 104 and 217. Three cases involved military planes stolen and crashed by unstable pilots, but of the rest, four were commercial flights whose pilots or copilots managed to crash their plane and take the passengers with them.

Jenkins resurrects the case of a Japanese pilot who crashed hid DC-8 into Tokyo Bay in 1982. He cites the case of the Air Force pilot who crashed his A-10 into a Colorado mountain in 1997. He states what so far nobody else has been willing to say, namely that “last March’s disappearance of Malaysia Airlines 370 appears to have been a criminal act by a member of the crew, though no wreckage has been recovered.”

The possibility of human error and human criminal actions is eliminated when the human element is removed. That is the clincher – if one were needed – in the case for free flight to replace our present antiquated system of air-traffic organization and control.

The case for free flight is analogous to the case for free markets and against the system of central planning and government regulation.

What if… 

Holman Jenkins reveals that as long ago as 1993 (!) no less a personage than Al Gore (!!) unveiled a proposal to partially privatize the air-traffic control system. This would have paved the way for free flight and automation to take over. As Jenkins observes retrospectively, “there likely would have been no 9/11. There would have been no Helios 522, which ran out of fuel and crashed in 2005 when its crew was incapacitated. There would have been no MH 370, no Germanwings 9525.” He is omitting the spillover effects on private aviation, such as the accident that claimed the life of golfer Payne Stewart.

The biggest “what if” of all is the effect on self-driving cars. Jenkins may be the most prominent skeptic about the feasibility – both technical and economic – of autonomous vehicles in the near term. But he is honest enough to acknowledge the truth. “Today we’d have decades of experience with autonomous planes to inform our thinking about autonomous cars. And disasters like the intentional crashing of the Germanwings plane would be hard to conceive of.”

What actually happened was that Gore’s proposal was poured through the legislative and regulatory cheesecloth. What emerged was funding to “study” it within the FAA – a guaranteed ticket to the cemetery. As long as commercial demand for air travel was increasing, pressure on the agency to do something about travel delays and the strain on airport capacity kept the idea alive. But after 9/11, the volume of air travel plummeted for years and the FAA was able to keep the lid on reform by patching up the aging, rickety structure.

And pilots continued to err. On very, very rare occasions, they continued to murder. Passengers continued to die. The air-traveling public continued to fume about delays. As always, they continued to blame the airlines instead of placing blame where it belonged – on the federal government. Now air travel is projected to more-than-double by 2030. How long will we continue to indulge the fantasy of government regulation as protector and savior?

Free markets solve problems because their participants can only achieve their aims by solving the problems of their customers. Governments perpetuate problems because the aims of politicians, bureaucrats and government employees are served by the existence of problems, not by their solution.