Why aviation innovation matters
Uber’s Elevate_ concept._
Aviation is not a big industry. According to the BEA, all of “transportation and warehousing” accounts for only 3 percent of GDP, and of course aviation is only a portion of that category. Despite the industry’s small economic significance in this respect, I’ve come to believe that working on aviation policy is one of the most valuable uses of my time as a policy researcher. Consider this post a defense of my decision to double down in this area for the next year.
What aviation looks like today
When most people think about aviation, they think mainly about commercial jet transportation. They are not wrong to do so. Large passenger airlines and air freight companies account for most of the value the industry produces. There is of course marginally more to it than that—there are helicopters and private planes and a nascent, limited commercial drone industry. But by comparison, these are small potatoes.
Today’s commercial jet transportation has a few peculiar characteristics worth noting. First, it is organized along a hub-and-spoke model. If you want to fly from a small city on the East Coast to a small city on the West Coast, you will almost certainly have to connect through a major hub.
Second, despite the fact that technology exists to fly supersonically, all commercial flights today are subsonic. The Concorde retired in 2003. No new supersonic commercial planes have entered operation.
Airplane Speeds Have Stagnated for 40 Years
Third, commercial transportation generally operates along a “common carriage” model. There are published routes and timetables; you buy a seat onboard a flight traveling a fixed route. If there is a sudden surge of demand of people wanting to fly a particular route, the system can’t really accommodate that.
Fourth, flights are still human-piloted. In fact, autopilot does most of the work, but there are still people behind the control stick who are ready to take control in the event of an emergency. Or, as in the case of Air France flight 447, to aggravate the emergency.
Fifth, air traffic control is still done by human beings. In the United States, these human beings still work for the government. US airspace lags that of other countries who have privatized their air traffic control operations, like Canada, the UK, Australia, and New Zealand. Raymond Russell has a great piece on this in Plain Text.
Sixth, aircraft designs are stagnant. Here is one of the ways that James Fallows describes the stagnation in general aviation in his book, Free Flight:
The most popular small [general aviation] planes, the high-wing Cessna 172 Skyhawk and 182 Skylane, are basically unchanged from the models first introduced in the early fifties. In 1997 the Beech Bonanza celebrated its fiftieth anniversary in more or less continuous production. Of course there have been changes over the years, mainly in instrumentation. But to see one of these planes is to have no doubt that it is an ambassador from a much earlier age.
It is possible to view such a venerable fleet as a sign that these were products built to last. This was the spin that an executive of the New Piper Aircraft company put on the situation in the nineties: “Our airplanes are well designed and well built, often remaining in service for thirty years or longer.” But almost any product — a car, a refrigerator, a 1981 original IBM Personal Computer — could also be kept going for decades, just like an airplane, if like airplanes it were subject to the federal requirement for top-to-bottom annual inspections, renovations, and repairs.
The only parts of the world where people actually try to keep cars going for decades, except when consciously preserving museum pieces, are places like modern Havana, where there’s no incoming supply.
The reason it’s not done elsewhere is that new products are better. With enough effort, you could keep that 1977 Gremlin running — but you wouldn’t, since the new models look better, are safer, run more efficiently, have new features, and are easier to maintain.
JFK disembarks from Air Force One in 1962 or 1963. It looks just like today’s planes, right? It flew at the same speeds, too.
Finally, between being forced to use overcrowded major airports and suffering the indignities of TSA inspection, the experience of air travel is fairly horrible. Many of us are pretty good at working within the current system—I have gotten from my front door to my airport gate in 16 minutes—but there is significant room for improvement.
How aviation could be better
The discussion so far suggests several margins on which the aviation industry could innovate. First and foremost, our uses of airspace aren’t scratching the surface of what’s possible. Airspace is a dramatically underused platform for innovation. In addition to humans ferrying humans and cargo around for long distances on giant tubes, we need (at least) drone delivery, 24/7 drone sensing and broadcast platforms, and “flying cars,” or in Uber’s words, “on-demand urban air transportation.”
But even regular air transportation has a lot of room for improvement. It could be made better with an emphasis on point-to-point travel instead of today’s spoke-and-hub model. Imagine countless direct flights on smaller planes using more convenient airports, with routes coordinated by software. Open your app, tell it where you want to go, and a service matches you with any other passengers traveling to the same city and a plane that’s exactly the right size. Of course, to make this vision scale and be affordable, the flights would need to be autonomous—there aren’t enough pilots in the country to fly all the point-to-point routes Americans would demand.
Human air transportation should be supersonic. Imagine flying to London in 3 hours instead of 7, to Tokyo in 6 hours instead of 11, or to Sydney in 7 hours instead of 15. That is all possible using known technology—Boom is trying to do it. If their Mach 2.2 jet were allowed to fly over land, you could cross the US in under 2 hours as well.
And, to be a bit greedy, Mach 2.2 is only the beginning. NASA and other science agencies are studying hypersonic (greater than Mach 5) concepts that would make anywhere on Earth accessible in under 4 hours. (Gratuitous sidebar: I like laser propulsion.)
To coordinate heavier and more heterogeneous use of our airspace, we would need smarter, more automated, and ultimately decentralized air traffic management. Aircraft should be able to communicate and coordinate separation with each other, whether they are manned or unmanned vehicles. Better air traffic management pays dividends with lower operational cost and more intensive use of the airspace.
And finally, we need to find a way to bring more dynamism to aircraft engineering. Cars today are radically better than cars of the 1960s. Why are aircraft not similarly improved? If aircraft had experienced the kind of transformation that autos have in the last half-century, we would be having a different conversation right now.
The bottom line is that flight of all kinds could become cheaper, faster, and more convenient. These improvements would have real benefits in terms of quality of life. Three times in the last two months I have taken 24-hour business trips that entailed a night away from home and my family (Austin, Denver, Waterloo). If aviation were better in the ways I am discussing, I could easily have made these into same-day trips. And with faster, cheaper same-day trip availability, I would be more willing and able to meet and talk to more people.
Imagine the benefits of being able to commute a much longer distance in less time through the air. Labor markets would become more efficient as we would all have wider job search radiuses.
Finally, as aviation brought the world together, more people would have the opportunity for significant international travel. Instead of going to Chinatown, more Americans could visit China. The benefits of these foreign cultural experiences are possibly substantial, providing a broader perspective for average people.
How do we get there?
To be sure, there are some technological hurdles to overcome before this vision of next-generation aviation comes to fruition. But a large percentage of the hurdles are regulatory. Here are some of the necessary policy reforms as I see them.
First, air traffic control privatization is probably necessary if we are to experience any significant modernization. As discussed above, other countries have privatized their air traffic management systems with great success. A proposal to privatize US airspace passed the House Transportation and Infrastructure last year but stalled because the Senate version of FAA reauthorization didn’t include it. The issue is almost certain to return in 2017.
Second, commercial drone operation is limited by regulations more than by the technology. The FAA is gradually relaxing the restrictions on drones by first granting waivers to certain elements of the rules and then slowly rolling the formal rules back. This process is likely to take years if the current approach isn’t challenged by Congress, as it should be.
Third, there needs to be a regulatory framework in place for autonomous flight. The drone portion of this is relatively straightforward. The hard part is rules for autonomous aircraft carrying human cargo (i.e., flying cars). The FAA should begin right now to draft rules, create test sites, and work with innovators to find out what kind of guidance they need. Flying cars seem like a fantasy, but we already have a pretty good idea in principle of how they might work. There aren’t any massive missing pieces. Consequently, now is the time to start working on a regulatory framework that would allow them to operate.
The Joby S2
Fourth, we should legalize supersonic flight over land. As Samuel Hammond and I explain in our recent paper, “Make America Boom Again,” the ban on supersonic flight should be replaced with a sensible noise standard for sonic booms. In addition, the FAA should use its leadership role at ICAO to push for economical takeoff and boom noise standards globally, so that supersonic R&D is able to have the benefit of a global market.
Fifth, aircraft type certification needs significant reform. It takes years for an aircraft to work its way through the FAA certification process. In part, this is because the FAA insists on using government employees to do all the certifying. In contrast, NHTSA, the automobile regulator, allows sophisticated car companies to self-certify that they meet regulatory requirements. This self-certification is combined with more direct agency oversight when cars reach the market. As a result of this less restrictive regulatory model, cars have experienced dramatically more improvement than aircraft have in the last 50 years. FAA should immediately move to the NHTSA model—they already have legislative authority to do so, but they just simply have decided not to.
In the long run, there are other models of ensuring that aircraft are adequately safe without requiring detailed regulatory oversight. One possibility that I want to explore is that of requiring the manufacturer to carry a large amount of insurance for its products. If this requirement were in place, insurers could substitute for regulators in a competitive, decentralized way to ensure that aircraft designs are efficiently safe. I think this sort of policy would lead to a flourishing of aviation engineering relative to the status quo.
Sixth, the FAA should allow greater experimentation for kinds of air transportation other than traditional airlines. Unfortunately, the FAA currently considers any air transportation that you can book online to be a common carrier, regulated in the same, heavy-handed way the airlines are. This has limited experimentation with forms of “flight sharing,” taking rides with private pilots who need hours in their plane to keep up their licenses. A case involving FlyteNow, a flight sharing service the FAA shut down, may soon be taken up by the Supreme Court. While I hope the Supreme Court rules for FlyteNow, nothing is stopping the FAA from lightening up on its common carriage definition on its own, and nothing stops Congress from defining common carriage less restrictively.
Seventh, it’s past time to legalize what is known as cabotage. You know how Europe has super cheap regional airlines that use smaller airports? What if they decided to expand to the US market? Wouldn’t that be cool? Turns out, they can’t, because it’s illegal. The US (and many other countries) ban foreign carriers from flying domestic routes. This is a strictly protectionist measure that enriches corporate interests at the expense of consumers. It’s ridiculous, and the ban should be repealed.
There are many more aviation policy issues that I probably won’t have time to get to in the next year. Alex Tabarrok had a great piece on Marginal Revolution about airport privatization, which is a great idea. Also, TSA reform or repeal is probably necessary if we want to make flying a truly pleasant experience. The Department of Transportation is considering rules for in-flight cellphone use.
Even with this limited subset of issues that I’ve highlighted, I have my work cut out for me. There are many billions of dollars in economic welfare on the line here, and almost no one outside of the aviation industry is working on these policy issues. The industry is hampered by their need to play nice with their regulator, so they are not the most ready messenger for the obvious conclusion that we need a fairly significant regulatory overhaul. My goal is to produce good evidence for most or all of these reforms by the time Congress reauthorizes the FAA in September. Wish me luck, or if you are a public interest policy wonk, join me.