Is the Air Traffic Control System Broken?

Kenneth LEVIN: When you’re sitting at the gate, and you’re getting ready to board a plane, do you have that moment where you think, “I’m putting my life in someone else’s hands?”

Stephen DUBNER: I do. This is just a habit of mine. Because my oldest brother was an Air Force pilot, and I’ve always been in awe, because I have none of the technical, mechanical, or optic skills, reaction time — none, I have none of those abilities. 

LEVIN: I’m with you. I’ve been around the industry for 30 years and still, when I sit at the gate or whatever, I’m like, Okay I know the team. I’m putting my life on their hands, and I’m trusting them, and that still goes through my head. It’s not that I’m nervous, it’s just that it’s what we do. people think about the pilots. Some people see the control tower, they think about us a little bit, but we’re probably, like, not that high up in the thought process. 

That is Kenneth Levin. He lives in northern California, and he recently retired after a 23-year career as an air traffic controller. He also happens to be a Freakonomics Radio listener, and a few months ago he sent us an email. He got right to the point. “Stephen,” he wrote, “How about an episode about Air Traffic Control?” My first thought was, “Yes, please!” A few years ago we did a three-part series about the airline industry, called “Freakonomics Radio Takes to the Skies,” but even in three episodes we barely touched on air traffic control. Like Levin says, it’s just not “that high up in the thought process.” I have since come to realize that that is by design. At any given moment, there are around 5,000 planes in the sky above the United States. The pilots of those planes are in contact with a constellation of controllers who guide the planes through departure, manage the separation between planes in the air, coordinate handoffs across the different 3D sectors in the sky — and, of course, there’s the landing. In a given day, there are around 45,000 flights in the U.S., and more than 3 million passengers. That’s about twice as many passengers as there were in the 1980s. It is a remarkably vast and complicated system — and it is also remarkably safe. If you look at commercial air travel globally, the risk of a fatality is around 1 in every 13 million passenger boardings. And that’s around 40 times safer than airline travel was just 50 years ago. The pilot is of course a high-profile part of this operation. But the people who control the air traffic — there are around 14,000 of them in the U.S. — they are hidden from sight. They are a sort of invisible hand, ensuring our safe passage. Although if something goes wrong, they stop being invisible. And lately, some things have gone wrong. In January, an Army helicopter and a passenger jet collided in midair just outside of Washington, D.C.; sixty-seven people were killed. A disaster of that scale can’t happen without a cascade of errors; but a top official of the F.A.A., the Federal Aviation Administration, did recently testify that errors by air traffic controllers may have contributed. And then in June, there were equipment outages at the control center in charge of Newark Liberty Airport; for stretches of 30 to 90 seconds, the controllers were not able to communicate with their planes. One of the controllers later described the event as “pure insanity”; some took trauma leave. Between these Newark outages and the D.C. crash, a certain amount of panic set in. Passengers avoided Newark airport. And for weeks, we all read about how outdated the U.S. air traffic control system is. But the panic began to fade as things returned to normal. And so, today on Freakonomics Radio, what is normal when it comes to air traffic control? And if normal isn’t good enough — which pretty much everyone agrees it’s not — what’s the best way forward? These are some of the questions we’ll be asking in this episode, and in next week’s episode too.

John STRONG: We basically have an iPhone 5 when the rest of the world has an iPhone 18.

We’ll hear the case for moving air traffic control out of its current government home:

Dorothy ROBYN: This is effectively a business trapped in a regulatory agency and micromanaged by Congress. 

And we’ll hear about a big injection of cash from President Trump’s One Big Beautiful Bill:

Ed BASTIAN: $12 billion is more than the F.A.A. has been allocated in years. So let’s get it going. 

Okay, let’s — because when you’re putting your life in someone else’s hands, it’s good to know how those people do their work.

*      *      *

Once you dig into the workings of the F.A.A., you quickly learn that it serves many constituencies. Sometimes their interests align, and sometimes they don’t. Let’s start with one big stakeholder: the commercial airlines.

BASTIAN: Ed Bastian, I’m the C.E.O. of Delta Airlines. 

Bastian and I spoke in July. We also spoke a few years ago for our earlier airline series, and we visited Delta headquarters in Atlanta. We toured their operations center, their training facilities, their meteorology department. Ed Bastian is coming up on 10 years as C.E.O., and Delta is the biggest airline in the world by revenue; it operates more than one-and-a-half million flights a year. An airline like Delta works with the F.A.A. in two distinct ways.

BASTIAN: They are a critical business partner, and they’re our regulator.

DUBNER: How would you describe the relationship between Delta and the F.A.A.?

BASTIAN: I’d say it was okay. It’s fine. It’s a tough challenge that they’re up against. They have not had the capital necessary. They haven’t had the resources allocated. They haven’t had the support from our legislators and our leaders in D.C. to execute with the level of quality that I think they would like to deliver to us.

DUBNER: How long has this been a business partnership that needs improvement? And how much do you think you’ve been able to advance the ball? 

BASTIAN: Over the last 20 years, it’s become increasingly clear, the need for investment and modernization. The reality is that these are systems that were developed in the 1950s and 1960s. Many of the same pieces of equipment from that era are still being deployed to this day. It feels like something out of a Star Wars movie gone bad. It’s not the F.A.A.’s fault. I mean, clearly it’s a tough job they have. But when you have a legislature that wants to govern — but only provide the necessary capital allocations on an annual basis or biannual basis, whatever the legislative authority at the time is, it has to be approved by Congress — it gets caught up in the political cycle. The nature of the investment and the need has been built up for such a period of time that it’s massive. 

DUBNER: You and I probably wouldn’t be having this conversation if there hadn’t been that fatal collision in D.C. and then a bunch of near- misses and other problems at Newark, and so on. Tell me how you as an airline think about passenger concern over safety. I’m also curious to know how your customers think about delays, and whether air traffic delays — especially during the summer — constrain demand from consumers who just say, Well, if it’s going to be that bad, forget it. 

BASTIAN: The accident at D.C. was shocking to all of us. We haven’t had an incident of that magnitude in 25 years in commercial aviation. So on the one hand, it does speak to how safe it is. We as an industry, commercially, fly three million customers a day, every day, and we do it relatively flawlessly. We execute, certainly from a safety concept, and customers over time take it for granted. We don’t want our customers worrying about it or thinking about it, because that’s what we do all day long. Safety is always the first and foremost matter of concern. We as an industry don’t talk about safety, in terms of comparing, “We’re safer than the other,” because we don’t compete on safety. When it comes to safety, we all learn from each other, we all support each other, we all do whatever we can, and as a result you have an industry where you’ve seen the safety results. We have that same relationship with the F.A.A. — open kimono. This is not cops and robbers. Everything we know, everything we see, we pass along to our regulator and vice versa. It’s only when you have full transparency of information and knowledge-sharing where you can actually start to create an environment where customers do look at safety and take it for granted. So we’ve had to educate more, I’d say this year, Stephen, on that regard. But now we’re not seeing any lingering effect from that incident now, six months later. 

DUBNER: Meaning demand — demand is still strong, you’re saying? 

BASTIAN: Demand is healthy. Now, with respect to delays, absolutely — the congestion in the sky, the storms that we’ve seen, a really difficult summer storm pattern. At Delta, we compare this summer to a summer a year ago. Storm activity of a disruptive nature is up 40 percent, that’s in the Southeast, that’s in New York, that’s on the West Coast, the middle parts of our country. It certainly seems to be trending in the manner there’s more convective activity, all of which puts more pressure on the system. What the F.A.A. does, and what we all do, is, everything slows down, and so customers are affected.

DUBNER: And how would you describe the state of the technology in the air-traffic control system? I’ve read, for instance, that the U.S. is lagging other countries in the use of satellite navigation versus the older radar navigation? 

BASTIAN: I’d say that we — you and I both have more technology in our car than we’re able to use in our planes.  

DUBNER: So, why, and what’s the way forward there?

BASTIAN: Well, why? It’s a radar point-and-shoot technology dating back to the ‘50s and the ‘60s. When you look in the sky, there’s a lot of lanes in the sky that are unutilized. Up in the northeast, you hear about storms in western Pennsylvania somehow keeping you, on a bright, sunny day locked down in LaGuardia, and it gets to be a credibility issue. Customers come to our people and they think we’re lying somehow. But there’s only a handful of pathways out. Runways are not built, pattern-wise, to deal with the storm. I mean, the LaGuardia runway patterns were created 60, 70 years ago. No one envisioned the level of congestion now.

DUBNER: Delta has spent billions of dollars remaking a LaGuardia terminal and a J.F.K. terminal — for which I thank you, because I happen to fly out of both those airports a lot. I have to say, the first time I landed at the new Delta LaGuardia terminal, I thought I was in the wrong city.

BASTIAN: That’s not uncommon, by the way. 

DUBNER: I almost got back on the plane and tried to go home to New York. But I’m curious how that investment seems to be paying off for you, especially with delays. And I know that that’s a pretty narrow airspace into LaGuardia already. So I’m just curious whether air traffic control or F.A.A. issues are affecting your ability to maximize your investment there. 

BASTIAN: It has. As an industry, we asked the F.A.A. several years ago to pull down the overall flying activity by 10 percent. So we’ve all voluntarily reduced our schedules, because you can’t get one airline to agree and the others don’t. We all have to work in concert. So by definition, there’s 10 percent less throughput in the sky than is authorized under the capacity studies that the F.A.A. legislates. A lot of this comes back to staffing too. This is not just about inefficient communications and technology. The F.A.A. is understaffed controllers by thousands. We have to schedule more time to fly to LaGuardia than we did in the 1950s when we launched that flight.

DUBNER: That’s your scheduled time, but from takeoff to landing, I assume it’s faster. 

BASTIAN: Capability-wise of course, we could go quickly, but what we have to actually allocate to ensure that we know what we can commit to our customers — yes, it’s a longer schedule than it was back in the 1950s.

DUBNER: How much of that added cost do you attribute to A.T.C. or F.A.A. uncertainty or complication?

BASTIAN: It’s hard to put a specific allocation on, but it’s considerable.

DUBNER: When I hear you describe all these problems or inefficiencies, it’s kind of remarkable to me that things work as well as they do. So is this a case maybe where we’ve got a system that, yes, is full of antiquated technology, but people find a way to make it work, and we should be grateful that it is working? Or is it more like, we’ve gotten away with it ‘til now, and it’s time to upgrade or there will be big trouble in the future? 

BASTIAN: It’s the latter, clearly. We know it well. The controllers do a phenomenal job with what they have. The airlines do a great job. We make the necessary accommodations. We slow things down as necessary. We pause whenever in doubt. The system will not last forever. Every day that goes by, we’re getting one day closer to when the thing will break.

So that’s how the air-traffic control system looks from the perspective of an airline C.E.O. — aging technology, staffing shortages, lack of investment. What does it look like from the regulator side? We couldn’t get an interview with senior officials at the Federal Aviation Administration — or at the Department of Transportation, which oversees the F.A.A. But we got a very good substitute.

Polly TROTTENBERG: Polly Trottenberg, former deputy secretary of the U.S. Department of Transportation, and I served for a time in 2023 as the acting Federal Aviation administrator. 

DUBNER: Thank you so much for being here. I really appreciate it. This is I think a really useful and interesting story for us to tell, which is really just how A.T.C. and the F.A.A. work, honestly. A lot of people just don’t know, so — 

TROTTENBERG: I — I would put it differently, almost nobody knows.

The Federal Aviation Administration started out as the Federal Aviation Agency, which was established in 1958 after a midair crash over the Grand Canyon a couple years earlier. It killed 128 people. In the aviation world, this is called “tombstone regulation” — creating new policy after a disaster — and which today the F.A.A. tries to avoid. There was, of course, plenty of air traffic control before the agency was founded. In the 1920s, farmers were paid to light bonfires across the Midwest to guide nighttime postal flights. Those were soon replaced by electric-light beacons, and radio followed in the 1930s. A system very much like the one we still have today — the A.T.C. towers and command centers, the radar technology — that came about after World War II, and it was built around the methods and technologies used during the war. Here’s Polly Trottenberg again:

TROTTENBERG: My father was a bomb dropper in World War II. He was a bombardier navigator. My father likes to say he was in the first plane that flew over Normandy on D-Day. I’ve not been able to confirm it. 

DUBNER: You distrust your father? 

TROTTENBERG: Well, who knows? Maybe many fathers say they were in the first plane that flew over Normandy on D-Day. My father was a very small man. He wanted to be a pilot, but they told him he was too short. They made him a bombardier navigator. He flew in B-26s. I got to climb into a B-25, which is a very similar plane, and I really understood it viscerally then, you have to go in your back through sort of a narrow duct to get into the nose of the plane. My father was very lucky. He had grown up very poor, no hope of higher education like so many after the war. My father got to go to Harvard on the G.I. Bill, and it transformed his life. 

DUBNER: Okay, so for someone who’s never thought about, maybe never even heard of the F.A.A., just describe how it works. 

TROTTENBERG: I’ll give an example that I think helps visualize the challenge the F.A.A. faces. F.A.A. runs a command center out in Virginia. It’s an extraordinary place to visit. You come in there, and you’ll see a radar map that is showing every flight that is flying over the continental United States. It’s an eye-popping number. They have enormous weather maps, and you can see in the continental United States, there’s always weather somewhere that they’re managing. There’s another radar screen that’s showing all the military operations, which we’re always doing a lot of. And then you go upstairs and there’s a room called the Challenger Room, that is showing what’s happening with commercial space launches — some of the most cutting-edge technology in the aviation sector. 

DUBNER: Also under the F.A.A. purview entirely, correct? 

TROTTENBERG: Correct. Just an astonishing range of things that the F.A.A. is following, monitoring as it’s guiding air traffic and just the range of technology from old, analog equipment to incredibly sophisticated commercial space launches. The F.A.A., in my opinion, is the largest, most complex, 24/7 safety-critical operation in the federal government. 

DUBNER: More than the F.B.I.? 

TROTTENBERG: Look, the F.B.I. is a big entity, but in any given moment there are thousands of planes in the air, hundreds of thousands of souls. You’re coordinating an operation of every airport around the country, coordinating with international aviation groups. It’s an enormous, complicated operation. 

DUBNER: And there’s never an hour off, a day off? 

TROTTENBERG: Correct. Except for some tremendously disruptive times. 

DUBNER: 9/11, you close the airspace and ground the planes. 

TROTTENBERG: 9/11, and then admittedly during the Biden administration, we had a problem with something known as the NOTAM system — Notice to Air Missions — where we had a technological problem and did have to shut the system down for a brief amount of time.

That shutdown led to more than 10,000 canceled and delayed flights — which is a lot of flights but, to keep it in perspective — remember, there are 45,000 flights a day in the U.S. The NOTAM system sends pilots real-time alerts about hazards or changes that may affect their flight, like a closed runway or an equipment problem. This NOTAM outage happened in early 2023.

TROTTENBERG: It was a pretty impactful moment and drove home a couple of interesting things. One, and I think it’s been obviously much discussed in recent months, that the F.A.A. infrastructure and operations are aging, and in desperate need of investment. Systems are starting to break down. Also just that, you know, the F.A.A. has so many systems. This one, the Notice to Air Missions, was not necessarily one people would have said was a safety-critical operation, but once you didn’t have it — pilots need that key information as they start to fly. So, it turned out it was a pretty critical system when it wasn’t operating properly. 

DUBNER: I’ve been reading about the shortage of air traffic controllers. I know there was a big push when you were in office to hire more controllers. So, tell me how significant the shortage is, and what can or should be done about it.

TROTTENBERG: There is a shortage, I think most folks would agree. There are a few factors at play. One is, you know, back in the mist of time, you may remember then-President Ronald Reagan got in a big conflict with what was then the Air Traffic Controllers union, and fired a bunch of them. 

This was in 1981. The controllers wanted better pay and shorter hours. Reagan, you may recall, years before he was president of the United States, had been president of the Screen Actors Guild.

Ronald REAGAN: Let me make one thing plain. I respect the right of workers in the private sector to strike. Indeed, as president of my own union, I led the first strike ever called by that union. But we cannot compare labor-management relations in the private sector with government. Government cannot close down the assembly line.

Reagan figured he could fire the strikers if he came up with enough replacements.

REAGAN: It is for this reason that I must tell those who failed to report for duty this morning, they are in violation of the law and if they do not report for work within 48 hours, they have forfeited their jobs and will be terminated. 

Reagan kept his word, and he fired 11,000 air traffic controllers. A new generation got hired in the aftermath; many of them later retired around the same time — and now, a second wave of controllers are again approaching retirement age. And there aren’t enough new recruits to replace them.

TROTTENBERG: It’s an incredibly sophisticated training process. And look, it’s a very difficult job. Credit to the men and women who do it. It is a job that requires some of the highest levels of skill. And it takes a long time to train them. During Covid, for a while, they weren’t training. And it’s a system where, even if you stop training for six months or a year, it takes a long time to get caught up. And then it’s just true, the system is growing and becoming more complicated, and you need more controllers to help manage it. And look, I think we’re in a — for better or for worse — a pretty transitional moment in the federal government. I worked in the U.S. Senate for 12 years, so I know from whence I speak. There’s always been a natural tendency — and this is just how our political system works, I don’t know that anyone’s to blame for it — that members protect federal facilities and employees within their states and districts. It’s sort of, almost political malpractice not to.

DUBNER: Although economists would argue it’s economic malpractice to do that.

TROTTENBERG: Well, economists — yes, occasionally they get elected to Congress, but not too often. 

Indeed, that is true — economists rarely get elected to Congress. But they do produce some relevant research: 

John STRONG: My work suggests that just on air traffic control staffing shortages alone, are like $135 to $160 million cost per year. 

That’s coming up, after the break. I’m Stephen Dubner, this is Freakonomics Radio. We’ll be right back.

*      *      *

You’ve probably heard this saying: “Two things can be true at once.” That’s a nice way to describe the U.S. system of air traffic control. It is both exceptionally safe and under a great deal of strain.

STRONG: Aviation is cheap, and it’s remarkably safe. But the salience of the last six to nine months has really been in the public view. 

That is John Strong. He’s an economist at the College of William and Mary’s business school.

STRONG: I’ve been working on transportation and aviation since the days of deregulation in the early 1980s.

DUBNER: And what got you interested in the first place in transportation generally, but aviation particularly?

STRONG: When I was starting into graduate school was the period at which all of the transport modes — trucking, air, rail — were all being deregulated. There was lots of change in the industry, and there was a whole wave of privatizations going around. So it was a – a really exciting industry to get involved in when you were a young graduate student.

DUBNER: And if you take those three — trucking, air, and rail — and the de-reg that happened there, how did air travel come out differently than the others?

STRONG: Deregulation meant that airfares came down dramatically. We had a country where less than 50 percent of the population had ever flown. And now it’s more along the lines of 90 percent. We had an explosion of new air services, lots of development at airports. But one place that’s lagged behind has been the infrastructure, the air traffic control to support all that growth over 40 years. The industry and the folks who are operating across the sector have known about these problems for a long time and have been sort of frustrated by, “It’s not broke, so don’t fix it.” I think we’re now at the point where we’ve realized that we need to do something. It’s highly significant safety events that spur actions that we probably should have taken earlier. 

DUBNER: If you had to shorthand the reasons why the infrastructure and the entire organizational architecture around air travel has not kept up, what would you say? 

STRONG: The growth in services and passenger demand created lots of opportunities, which immediately affects airports, of course. Airports see their customers as airlines, so they’ve been quite responsive to that. Because air traffic control has traditionally been a government function, it’s been slower to respond to the changes in the industry conditions. For all the new technology that’s come into air traffic control, we’ve been woefully slow and we’re getting way behind in that technology investment in A.T.C.

DUBNER: It sounds like you’re saying there is a lot that could be done to make the system both more efficient and safer. Is that right? 

STRONG: I think that’s right. If you look at the larger plan that the F.A.A. has been working on for some time, what they call the Next Generation Air Traffic Control System, essentially shifts air traffic control from a ground-based radar system that we’ve had since the 1950s to a satellite-based navigation system that relies on GPS. The older model was a command-and-control kind of model, and now we have lots more communication and satellite-based capabilities that we’re just not taking advantage of.

DUBNER: I see. What would a sat-nav system cost?

STRONG: The estimates vary, and it depends how far and how fast you go with it. The rough idea of people talking on the order of $40 to $50 billion for the whole system. Many of these facilities are more than 40 years old. We’ve seen the problems at Newark in particular with respect to telecommunications as well as the radar and satellite-based nav. Think about how amazing it is when you travel in your car, to be able to use satellite-based navigation to get you wherever you want to go. We’re not taking advantage of that technology in the air the same way that we are in the ground. 

DUBNER: I would think that modern compute power would be a massive help in air traffic control, whether it’s A.I., machine learning, all these forms of automation that help humans do the jobs they were already doing, but do them better. What’s the status of — let’s just call it A.I. as a shorthand — within air traffic control? And I’m curious if there’s a reluctance to accept, where that reluctance is coming from? 

STRONG: The needed investments are not to automate, and remove controllers from the system. It’s to give them better information in a more timely fashion. And what the new technology really lets controllers do is manage air traffic rather than control air traffic, right? If you go into a tower and you look at the screens today versus the screens 15 years ago, it’s very, very different. There’s lots of current technology, which we need to roll out faster, which helps the controllers do their job more efficiently and more safely. The challenge of how much could we automate through A.I. is really a next-stage question. since this is a system that operates 24 hours a day, 7 days a week, there’s 5,400 planes in the air at peak, we can’t shut it down and switch over. If you look across the air traffic control facilities, 313 facilities that the F.A.A. runs, every one of them has a different kind of airspace, different traffic patterns, different weather patterns. And so the ability for A.I. to come and learn 313 different things, it takes time.

DUBNER: Maybe that helps explain what I wanted to ask you about next, which is this report that you recently wrote about staffing issues among air traffic controllers. You say that there are 19 understaffed facilities that account for about 40 percent of all delays, but then there are about 30 percent of facilities at levels that are more than 10 percent above their staffing targets. When I read that now, having heard what you just said — I used to think that you could just pick up air traffic controllers and move them from one location to the next. It sounds like that’s a lot harder to do than a civilian might think.

STRONG: Yeah. So everyone agrees we have a shortage of controllers. The number depends on who you ask, and what assumptions you make. The union says 3,000, the F.A.A. says about 1,000. But somewhere in that neighborhood. In order to produce more controllers for the facilities that are the busiest, it takes about four to five years. If you apply to be a controller, you pass your initial screening, which is physical and mental health tests and aptitude tests. Then you go through training. Then you essentially go out into an apprenticeship model — you go out into a facility, and you’re known as a developmental controller, where you’re learning how to operate the system in that space, right? Then there are full certified professional controllers who are capable of doing all of the operations in that facility. The problem is, suppose I wanted to move you from one facility that is overstaffed to a facility that is understaffed. When you move to that facility, even though you might have been a full controller in the old facility, you have to learn the new facility that you’re going to. So you become a controller in training for some period of time. The F.A.A. has not had enough flexibility to transfer folks from where we have overstaffing to understaffed facilities, but even when they do that, it takes time to fix that shortfall. 

Here is another factor that contributes to the shortfall of air traffic controllers: the U.S. has only one major training academy for controllers; it’s in Oklahoma City. Last year, an F.A.A. funding bill included plans for a second academy — maybe in Chicago, Dallas, or New York — but those plans were scrapped after opposition from lawmakers in Oklahoma. I asked John Strong how helpful it would be to have multiple A.T.C. training academies.

STRONG: In the long run, that may be helpful. In the short run, there are two things that have lots of potential to help. First of all, the academy in Oklahoma City has made some investments which are really going to help with training capacity. There’s more capacity there than we’ve been effectively using. The other dimension that’s important is that there are a number of programs across a number of universities, which does a lot of the basic air traffic control training. So folks wouldn’t have to go to Oklahoma City for that. They could do it at a number of these university partnerships and then go for final graduate level training at the academy. Which would get more folks into facilities for their actual on-site training and development faster. In the long run, a second facility might make sense. In the short run, we can take what we have now and use it better and faster. Right now, about 75 percent of the folks that come into training make it through. If we could get that number up to 85 or 90 percent, that would solve a lot of the problem.

DUBNER: Talk a bit more about who’s paying for the system. How does the F.A.A. raise money?

STRONG: The F.A.A. is part of the Department of Transportation. The total F.A.A. budget is about $18.5 billion. Eighty to 85 percent of that comes from distributions from a trust fund — which, every time that we fly, we pay a 7.5 percent ticket tax. There’s a waybill tax on cargo. And there’s a few other fees that go into that. And that creates a trust fund. The other 15 to 20 percent comes from the general fund. That’s appropriated by Congress each year. We have this mixed structure. If you take the total F.A.A. budget, about two-thirds of it goes to operations — about half of it goes to air traffic control, and about half of that is to pay controllers. So the amount that’s left for investment in equipment and facilities and infrastructure and new technology is pretty constrained. We spend about $4 to $5 billion a year on R&D and facilities and equipment in air traffic control. About 90 percent of that facilities and equipment budget is for repairs and maintenance on the system we have, not new investment.

DUBNER: What you’re describing, John, sounds like a pretty typical situation, which you often see in governments, but it — it happens elsewhere as well, where you spend money today for the things that you have to fix or keep up today, but the things that may deteriorate over time, or the things that you may want to have better years from now, you’re essentially ignoring. Is that too blunt a way to describe it? 

STRONG: You can always push it off until tomorrow. That’s been the mode of operation. Eventually, the equipment gets old. It’s just too hard to maintain. There’s lots of discussions about Windows 95 and paper flight strips. While there are a few legacy issues, there’s a lot of modernization that has been done at the F.A.A. They run a huge operation, 45,000 flights a day, just under three million passengers, and they do it remarkably safely. But the way that we maintain safety is at the cost of efficiency. We’ve all seen how airlines have buffered their schedules for flight times. Passengers do the same thing, buffering for connections. Airlines do the same thing, buffering for safety stocks of maintenance and crewing and staffing. It’s a hidden cost of the operation of the system. My work on delays suggests that just on air traffic control staffing shortages alone are like $135 to $160 million cost per year.

TROTTENBERG: There’s tremendous complexity to the budget process.

That, again, is Polly Trottenberg, who during the Biden administration was the number two official at the Department of Transportation, and for a brief time was acting head of the F.A.A.

TROTTENBERG: You have to step back and look at how the federal government budgets right now. I mean, talk about a system that has grown overly complicated and doesn’t work very well. The federal government here in the United States doesn’t pass budgets on time. Different parts of the budget have funds that can be called mandatory spending. What it basically means is, “Here, entity, agency, we give you this money, it’s guaranteed, it won’t be subject to the regular appropriation cycle, and you can spend it as you need it.” That kind of funding is particularly important for capital investments. For highway spending, for example, state D.O.T.s will know what kind of money they’re going to have for the next 10 years, so they can plan their capital budgets. Ironically, we give out a lot of money to airports around the country. That money is mandatory spending, so the airports know that they will have years to invest it, to improve their facilities. The F.A.A.’s major capital money, which is called the Facilities and Equipment Account, F&E, is subject to the discretionary budget caps in the federal process, which means you don’t necessarily know what you’re getting until Congress passes a bill which can often be halfway through the fiscal year.

DUBNER: I assume it also means that every single decision is subject to the perhaps idiosyncratic, let’s call it, whims of a single congressperson? 

TROTTENBERG: I think less that, and just that it’s competing against everything else in what is the pretty small and tightly squeezed part of the federal budget. The result for the F.A.A. historically has been that that Facilities and Equipment Account has been around $3 billion for around 15 years, not really even increasing for inflation. That is dramatically less than the agency needs to not only maintain this existing inventory of facilities and equipment, but to do something we’re all craving, which is modernize the latest technology.

DUBNER: So, would you say that in the U.S., air traffic control is in a crisis mode? 

TROTTENBERG: I don’t think it would be fair to say crisis, but I do think it’s in pretty urgent need of attention. Some of that attention is coming. So, there’s a bit of good news to the story. Not everything they need, but a pretty strong down payment. 

DUBNER: Let me switch gears for just a minute. I feel it’s an overlooked tragedy, or maybe just something we habituate to, which is how many people still die in this country, and elsewhere, from traffic crashes. Tell me how the federal work on airline safety has differed from the federal work on road safety, and what one might learn from the other.

TROTTENBERG: Thank you for this question. You know, there had been a big spike in roadway fatalities nationally in the Covid period. Over 43,000 people dying on roadways. Those numbers mercifully have started to come back down, but it’s an astonishing loss of life. Not just the people who die, but the people who have life-changing injuries. Hundreds of thousands of those, and the damage to families and communities is dramatic. Big difference between the two systems. The aviation system — the federal government, in a lot of ways, controls it from A to Z, from certifying pilots to certifying airplane designs to running the air traffic control system. It’s a contained, controlled, very high-level system. Our roadway system is a much more decentralized system, where any old 16-year-old can basically get a driver’s license, and the federal government has much less control over the designs of automobiles, of who gets to drive, of how local law enforcement handles traffic safety.

DUBNER: To be fair, We want everybody to be able to have individual mobility. I assume you’re not advocating for driving to go through the federal government.

TROTTENBERG: Absolutely not, but having worked on roadway safety at the local level in New York, there are a lot of things we can do to make systematic changes that make the system safer. Improving roadway designs, speed cameras, education, very targeted, sensible enforcement, better driver education, working with fleets who often set the tempo on the road — taxi, Uber and Lyft drivers, bus drivers, truck drivers. We have a lot of proven methods to make things safer on the roads, but they’re much more complicated and decentralized than an aviation system, where so much of it is controlled by a small number of entities. 

Coming up after the break: how to coordinate — and fund — that small number of entities:

ROBYN: The air traffic control issue became my white whale. 

Also, if you want to hear more about how complex systems need to be constantly upgraded, check out an episode we made years ago, No. 263, called “In Praise of Maintenance.” It is much more interesting than that title may imply. I’m Stephen Dubner, this is Freakonomics Radio, we’ll be right back.

*      *      *

Okay, so if you wanted to identify the main conflict that has kept the air traffic control system in the U.S. from keeping up with the times, what would it be?

ROBYN: You can’t run a business out of a regulatory agency. But I really would fault Congress. They should know better.

This is someone who has done a lot of work on the F.A.A. and air traffic control.

ROBYN: I’ve long thought it should be a Freakonomics episode.

DUBNER: Why didn’t you tell us? 

ROBYN: I should have. I should have. 

DUBNER: Why don’t you just say your name and what you do.

ROBYN: Sure. Dorothy Robyn. I’m a senior fellow with a tech-oriented think tank in Washington, D.C., I.T.I.F. I’m a policy wonk. I spent eight years in the Clinton White House on the economic team, and I was in the Obama administration at the Pentagon and the General Services Administration. I kind of specialize in those issues where it’s economists speaking truth to power. And they’re often very politically difficult. 

DUBNER: I read that you were special assistant to the president for economic policy at the N.E.C., and that you coordinated policy issues related to aviation, aerospace, defense, telecommunications, and science and technology. I’d love people to get a sense of how much goes on behind the scenes to make these things work that we often don’t really think about.

ROBYN: Being on the National Economic Council, was the eight most exciting years of my life. Clinton’s economic team was in the catbird seat in the sense that he was focused like a laser on the economy. So what we did mattered. Democrats had been out of office for 12 years. We had a lot of ideas about how to make things better. And it was a group of generally very market-oriented people. Now, there were divisions within. Bob Reich and Bob Rubin didn’t agree on everything. But I don’t remember ever feeling like I wasn’t philosophically comfortable with something. I had a very broad portfolio, as all White House staff do, and mine changed over time, so it sounds even broader when you read it. I picked up the aviation portfolio because of my transportation background.

DUBNER: How did you get to know everything you knew to be valuable in the Clinton and Obama administrations? 

ROBYN: I did an M.P.P. in public policy at Berkeley and then stuck around and did my Ph.D. there as well. Relevant to this topic, I wrote my dissertation on trucking deregulation, which came right on the heels of airline deregulation and involved many of the same players, and the politics were similar. I was a fly on the wall watching it happen. Economists had made the case for trucking deregulation. The question was, how could such a good idea get through Congress? Which it did, remarkably. 

DUBNER: What do you mean by that? 

ROBYN: Well, trucking deregulation was opposed by labor and the trucking industry. They had big stakes. They made a lot of money off of regulation. The beneficiaries would be consumers. So, diffuse benefits, concentrated losses — that is normally a formula for the consumers to lose. That was also true in airline deregulation. Economists had made the case for the consumer benefits, but there was industry and labor opposition. And remarkably, because of a variety of factors, including Jimmy Carter and a ragtag group of liberals and libertarians, both of these reforms got through Congress and laid the groundwork for a set of additional reforms of a similar nature.

DUBNER: So, airline deregulation was 1978, is that right?

ROBYN: Exactly. And trucking was ‘80. The Carter White House felt like, we’ll take on airlines first because they’re not as politically powerful as the trucking industry and the Teamsters. If we can get airline deregulation done, we’ll go after trucking. 

DUBNER: I didn’t know that — that’s an interesting strategic sequence. Why was the airline industry easier to — I don’t want to say push around, but easier to dereg?

ROBYN: There was no airline group comparable to the Teamsters. And the stakes were higher in trucking. You had a group of companies that had very valuable operating certificates, similar to taxi medallions in New York, the scarcity value. And deregulation would presumably eliminate that value. The airlines, they were comfortable in their regulated hot house, but they didn’t make a lot of money. And so, the stakes were not as high.

As we heard earlier from the economist John Strong, airline deregulation brought prices down. Those lower prices created a lot more demand for flying, and all that flying began to put a strain on the air traffic control system. It’s been a battle ever since to keep A.T.C. infrastructure up to date. I asked Dorothy Robyn how she thinks about deregulation all these years later, for both trucking and airlines.

ROBYN: I am very, very high on, on both of them. I actually don’t spend much time talking about trucking deregulation. There isn’t much controversy about that, except among certain liberal, lefty labor groups. Airline deregulations — because service declined, and it declined very predictably because load factors went up — it’s gotten a bad name in some respects. Consumer groups who supported deregulation — Ralph Nader, their support was absolutely critical. They tend now not to be supporters of it because of the quality of service, some degree of the impact on jobs. So, I’m a cheerleader for it. I think it has been tremendous for consumers. 

DUBNER: Are you sure you’re a Democrat?

ROBYN: I’m a — I’m a New Democrat. I’m a market-oriented Democrat.

DUBNER: What do you think has been a greater decline: service factor or, you know, prices and access to flying? I mean, how do you weigh the pros and the cons essentially of dereg?

ROBYN: I think airlines are giving consumers what they want, which is lower fares at the expense of service. My advice to people who complain about it is, fly business class. You will be paying about what you would have been paying had the airlines not been deregulated. The back of the plane is brought to you by airline deregulation.

DUBNER: Here’s something you wrote recently. “The ongoing crisis at Newark Liberty International Airport has laid bare the sorry state of our nation’s air traffic control system. Equipment outages have left air traffic controllers unable to track or communicate with planes for multiple 30- to 90-second intervals,” which I gather is quite a long time when you’re flying an airplane. “Some controllers,” you wrote, “have had to take leave to recover from the trauma, exacerbating a staff shortage and forcing the Federal Aviation Administration to further curb flights at Newark.” Okay, that sounds — I’m just going to say it, it sounds third world. It sounds poorly run. It sounds scary. It sounds frustrating. It sounds expensive. How did we get to this point? And I also don’t understand how we reconcile your argument that the F.A.A. and air traffic control are so outmoded with the reality that we still have very few commercial crashes.

ROBYN: When a crash occurs, a whole lot of things have gone wrong. It isn’t one single thing. Even Newark was, you could argue, that was somewhat exceptional because there was a closed runway. They had recently moved controllers from Long Island to Philadelphia to oversee flights in and out of Newark remotely. A lot of this is done remotely. And so, there were some additional technology challenges. But basically, my understanding is the copper wires were fried. There was fundamental technology there that did not work, and they didn’t have backup systems in place. And you are seeing more and more reports of that kind of thing happening.

DUBNER: Sean Duffy, the current Department of Transportation Secretary, in his new plan to modernize F.A.A. technology, there are photographs. One shows the hardware of an old radar system being cooled by a little desk fan. It shows paper flight strips instead of electronic. 

ROBYN: All true. All true.

DUBNER: There’s a current Alaska flight-planning system that shows ribbon cable and aluminum foil to reduce interference. I mean, it sounds absurd to me.

ROBYN: It is. It’s crazy. This is a piece of infrastructure that everybody else has figured out how to run better, in a way that allows it to borrow money like any company or state and local government can do, to make long-term investments. To charge users directly. I mean, it’s so obvious what the answer here is. It was obvious in 1995, and only four other countries had done this. Now, close to 100 countries have done this.

The obvious solution that Robyn is talking about has to do with the relationship between air traffic controllers and the federal government. During the Clinton Administration, Robyn was a part of an effort to get F.A.A. funding out of the Congressional appropriations process. This would have also separated air traffic control from the F.A.A.’s regulatory duties. A similar proposal came up during the first Trump administration. When Robyn says that nearly 100 other countries have by now “done this,” she’s talking about what she calls corporatization, although some people think of it as privatization.

ROBYN: Yeah, we said “corporatized” at the time. It’s a semantic thing. I just try to avoid the P word because people make that out to be, Oh, it will be run by a for-profit entity that will listen only to the airlines. This is a 24-7, high-tech “business,” in quotes, trapped in a regulatory agency, that agency being constrained by all kinds of stultifying federal rules and practices and micromanaged by Congress. It doesn’t require the kind of policy judgments or trade-offs that policing or national security does. Basically, it’s a production line. It’s complex, it’s safety critical, but it’s a very rule-based activity. Most countries that have spun off air traffic control — and it’s now all but a couple of developed countries, I think France and maybe Greece — most of them have made it a government corporation, which is what the Clinton administration proposed.

DUBNER: What’s a parallel government corporation that we’d be familiar with?

ROBYN: A better way to think about it is, it’s a utility. You have investor-owned utilities, municipally-owned utilities, and you have cooperatives. Most of the air traffic control systems are the equivalent of municipally owned. So the government still has some presence. They’re able to operate like a business. They are able to borrow money. We haven’t talked about borrowing authority. It’s absolutely critical. The Canadian model is the only co-op model that’s out there. The U.K. may be the only one that have a for-profit system, but most of them are government corporations.

DUBNER: Say a little bit more about NAV Canada, the A.T.C. provider there.

ROBYN: It’s been extremely well-run. It’s hard for me to tell how much of that is the people and how much is the model. I like the model because the incentives are aligned. The users want to have the most efficient system. You don’t need outside rate regulation like you do with utilities. It’s the users running the system. They have cut their costs and improved their technology to a remarkable degree.

DUBNER: Looking back at the countries who have corporatized or privatized their air traffic have you or anyone else ever empirically compared the outcomes of those versus those that remain regulated within the government?

ROBYN: Generally speaking, this has been a successful effort to get a business-like activity out of a traditional regulatory agency. Above all, for safety reasons, to get away from the inherent conflict of interest when you have the safety regulator also operating the air traffic control system. There’s been quite a bit of work done — John Strong and Clint Oster wrote a book about this, back in 2007.

That book is called Managing the Skies: Public Policy, Organization, and Financing of Air Traffic Management. Here, again, is John Strong, the transportation scholar at William & Mary, on the virtues of pulling air traffic control out of a government’s transportation agency. 

STRONG: It enables the air traffic control organization to be responsive to the different stakeholder communities, but probably even more importantly, it gives them the ability to raise capital and to make capital investments. The government has a controlling stake, but they operate with a separate capital budget. So, NAV Canada, it has a 15-member board with all the stakeholders represented, including the government, and what they’ve been able to do is create an organization in which they can manage capital and operations in a more flexible way than we do here with a structure in which all of the interested parties have a voice.

DUBNER: There have been multiple efforts to privatize, or let’s call it corporatize, air traffic control, and they’ve repeatedly failed in Congress. Why? Where is that pushback coming from? What are the political motivations there — political or maybe economic motivations?

STRONG: I think there’s probably three sources of why we’ve been slow to do that. The first is that in some ways, the F.A.A.’s real customer is Congress.

DUBNER: Is that because congresspeople fly a lot, or just because congresspeople are congresspeople? 

STRONG: Well, they fly a lot, but also a lot of them have F.A.A. facilities in their districts. They have towers and centers, and so they have a localized view of that. A second aspect is that the F.A.A. hasn’t been great over the years at managing capital investment, at managing upgrades and modernization, and so I’m reluctant to give them a lot of money because I don’t think it’s going to work. And then the third and most important reason is, Okay, I know we need to do something, but the system works, and I don’t know if a new system does. So if it’s not broken, don’t fix it. And if the cost of the old system is some more delays and some reduced efficiency, but it’s safe, then let’s just leave it alone. 

Okay, so those are some of the key reasons that the American air traffic control system is not as robust or modern as it ought to be. Dorothy Robyn has one more direction in which she’d like to cast some shade. 

ROBYN: In a recent piece I did in The Atlantic, I pinned it on general aviation, and that really refers to two different groups. There are private pilots — dentists who fly on weekends. And then there are business jets — C.E.O.s, companies use them, sports teams use them, the .1 percent. Those two groups, the first one is represented by a group called the A.O.P.A., Aircraft Owners and Pilots Association, the second by a group called National Business Aviation Association. The first group, there are a lot of them. Maybe 400,000 members. The G.A. caucus is perhaps the biggest caucus in Congress. They’re in every state. N.B.A.A. is a much smaller group. They perceive that they have something to lose because they pay almost nothing to use the current system. So they’re subsidized, a billion a year. The two efforts — in the mid-90s by President Clinton, in 2016 and ‘17, ‘18, led by the chairman of the House Transportation Committee, Bill Shuster, profile in courage in my view, Republican support from the Trump administration, although not enough support. Shuster modeled his legislation after the Canadian version, NAV Canada, because it has proven to be so effective. But what both efforts did was endeavor to hold harmless the general aviation groups who opposed it.

DUBNER: Let me just understand how the political pressure works. So, you said that there are something like 400,000 general aviation pilots in the country. I know a lot of people who fly these small planes, and I could imagine that they’re pretty good at banding together to at least contact their representatives to say, “This new proposal is terrible for us.” Is that the way it works?

ROBYN: Exactly. It’s very one-on-one. The private pilots, they’re not really a problem in terms of trying to move to a more efficient system, because they stick to small airports, they are not flying in an airspace that is controlled by air traffic controllers. They’re just afraid that a more efficient air traffic control system that’s going to somehow hurt them — a lot of federal money flows to airports — and that they’ll have less access. For the business jets, they fly in the same airspace.

DUBNER: And sometimes the same airports, yes?

ROBYN: Sometimes the same airports, yes. And they are paying a fraction of what a commercial airliner is paying for a similar trip. Let me use an example that the Eno Transportation Center, it’s a transportation think tank, used. This goes back to 2017. Two flights from Dallas to D.C. One is a passenger airline, 180 passengers, Airbus A21, and the other is a private jet with 10 passengers. According to this Eno example, the passengers pay $2,600 in the form of a ticket tax. There’s also another $760 in segment fee. That’s a flat charge. And then $100 in fuel taxes. So, a total of $3,500 in taxes. The private jet — there is no ticket tax. There’s no segment fee. All there is is a fuel tax of a couple of hundred dollars. In their example, $3,500 versus $327.

DUBNER: But to be fair, the big plane is carrying a lot more people. 

ROBYN: But here’s the key. A blip is a blip.

DUBNER: When you say blip, you mean the blip on the radar of the air traffic controllers?

ROBYN: Yeah.

DUBNER: A plane is a plane.

ROBYN: A plane is a plane. If you’re a controller, and you’ve got an A321 and a business jet, they get equal treatment. The cost they impose on the system is identical.

DUBNER: It sounds as though you’re saying that even if you represent only one percent of a market like this, that if you’re able to make enough noise, you can stall reform, is that right?

ROBYN: Exactly. I would add another reason, which is that trade associations, interest groups need issues. They need issues in order to demonstrate their worth. Sometimes they will take positions that may or may not be aligned with their members’ interests. A lot of those companies that own business jets, I got to believe those C.E.O.s would understand the benefits of corporatization of air traffic control. But N.B.A.A. has made this a very effective issue. 

DUBNER: So, it sounds like we should interview someone from the National Businesses Aviation Association?

ROBYN: Yes. Yes.

Until then, take care of yourself — and, if you can, someone else too.

  *      *      *

Freakonomics Radio is produced by Stitcher and Renbud Radio. This episode was produced by Theo Jacobs and edited by Ellen Frankman; it was mixed by Eleanor Osborne, with help from Jeremy Johnston. The Freakonomics Radio Network staff also includes Alina Kulman, Augusta Chapman, Dalvin Aboagye, Elsa Hernandez, Gabriel Roth, Greg Rippin, Jasmin Klinger, Morgan Levey, Sarah Lilley, and Zack Lapinski. Our theme song is “Mr. Fortune,” by the Hitchhikers; and our composer is Luis Guerra.