Gridlock at 30,000 Feet

Passengers at La Guardia on October 26's American Eagle Flight 822 to Des Moines. It was scheduled to depart at 5:55 p.m. but took off at 7:23 p.m.Photo: Vincent LaForet

It’s not your imagination: Flying into or out of New York is the worst it’s ever been. In fact, our airports are the worst in the country. According to the FAA, Kennedy, La Guardia, and Newark hold the three bottom spots in the national on-time arrival competition. Newark has the most delayed landings, with only 57 percent of flights coming in on time. La Guardia is second to last, with 58 percent, and JFK is third worst, with 59 percent. Leaving is no picnic, either. JFK boasts the most sluggish takeoffs of the nation’s 32 major airports: Barely three out of five departing flights leave the so-called Gateway to the World on time. (Newark and La Guardia don’t fare much better on that front, ranking No. 30 and No. 25, respectively.) All in all, over one third of the nation’s delays happen in the New York metro area. The Port Authority estimates that our airports’ inefficiency will cost us $200 million this year in passengers’ lost time alone.

But if delays began and ended in New York, we could probably file them under Colorful Local Annoyances, alongside subway woes and summer blackouts, and accept them as yet another levy for living here. When it comes to airports, however, New York’s problem is the world’s problem: Close to 3,700 flights stream through our three airports daily. And our sluggishness infects the entire grid. “If you look at the delay signature in New York and on the national level, you can see it propagating delays into the national system,” says John Hansman, an MIT scientist who studies air-traffic patterns. When a tardy Kennedy flight finally makes it to a hub like O’Hare—already stretched thin itself—air-traffic controllers are forced to put other, more-punctual arrivals in holding patterns to accommodate the one that’s already outrageously late; the ensuing domino effect can mean missed connections and lost luggage from Seattle to Sarasota. According to Kevin Mitchell of the Business Travel Coalition, 75 percent of delays around the country originate in New York. Since only 33 percent of the total delays actually happen here, this means that every local delay triggers more than two delays elsewhere.

The problem isn’t entirely new: Our airports have been watching their performance go down the tubes ever since the U.S. airline industry started digging itself out of its early-aughts slump. This year, however, we reached the breaking point. At JFK, delays beset 33 percent of total flights in the first quarter. By July, that number hit 41 percent. Extreme delays—one hour and over—are up 114 percent. Everybody’s heard about the system meltdown that kept hundreds of JetBlue passengers imprisoned on nine jets for six to ten hours this February at JFK. That may have been the most highly publicized instance of the dysfunctional system, but it was far from the only one. Six months later, in August, an American Airlines flight to La Guardia spent more than five hours glued to the tarmac. And a recent Continental flight from Caracas to Newark was diverted to Baltimore and stranded there for five hours. (Passengers were detained an additional two hours for questioning, after a few onboard voiced their impatience a little too forcefully.)

By September, New York’s airport mess was beginning to look a lot like a national crisis. Once the August delay figures came in—clinching the worst summer on record, here and nationwide—Washington started to pay attention. “Don’t be surprised when the government steps in,” said FAA chief Marion Blakey in a speech on September 11, just before leaving her post. Soon after, President Bush was grilling Blakey’s replacement, Bobby Sturgell, and Sturgell’s boss, Transportation Secretary Mary Peters, on the JFK situation. The result of the meeting was the creation of a quick-acting task force with two months to formulate a solution. “We’ve got a problem, we understand there’s a problem, and we’re going to address the problem,” declared Bush. The message was clear: We’ve got an emergency on our hands, and New York is at the epicenter.

Part of the problem, of course, is the sheer number of people moving through our packed runways, overloaded terminals, and crammed airspace. Last year, New York airports handled 104 million fliers—and nearly 1.25 million takeoffs and landings (which is more than Chicago’s O’Hare, the nation’s busiest single airport, combined with Midway). JFK handles five times as many international flights as any other airport in the U.S. and also happens to be the country’s premier destination for foreign freight. The peculiarity of our location—we’re practically equidistant from the West Coast and Western Europe—means that late afternoon, our rush hour for domestic departures, is also the peak time for international arrivals. Beyond that, there is the issue of massively increasing demand: If current growth holds up, by 2025, we’re expected to be processing 150 million travelers a year. The forces driving the flood of fliers include New York’s swelling population, a healthy economy that puts more travel money in consumers’ pockets, a recent boom in private and chartered aircraft, and, last but not least, the rise of the no-frills carriers. (The poster child for low-cost flying is JetBlue; the JFK-headquartered company has gone from zero to the largest single-airline presence at the airport in less than eight years.) “We just have too many aircraft in the system,” says Jim May, an airline lobbyist.

If only it were that simple. The fact is, it’s not just a question of having too much water in the pipes, so to speak. We also have too few pipes. And those that we have are rusting. All of New York’s airports, not to mention those in Philadelphia, D.C., Boston, and the rest of New England, use one traffic-choked corridor for the lion’s share of domestic flights, a kind of airborne I-80. (The route’s official FAA name is J80.) Over and around the city, the superhighway twists into a clump of weaving approach paths leading to the Big Three, as well as Teterboro, Stewart, Islip, and other airfields. Industry insiders like to compare the air-route tangle to a plate of spaghetti; a more accurate description might be a three-dimensional subway map. A few paths lie directly, one might say unnervingly, under one another. Departures from Teterboro’s Runway 19, for instance, can look up and see the undersides of jets headed for Newark’s Runway 22. (Teterboro jets often use a rule-bending maneuver called the Dalton Departure to avoid waiting their turn: They take off under a laxer protocol normally used by small planes, sneak under the Newark traffic, then file a proper flight plan in midair.) A few other paths are in intermittent conflict. Some routes date back to the times when pilots found New York by following the Hudson, aided only by bonfires and beacons; if you’ve ever wondered why your LAX–LGA flight treats you to a nice view of the Atlantic Ocean, followed by a stomach-testing 280-degree turn, you have to go back decades for the answer. The last tweaks to the system were made twenty years ago, and by the FAA’s own admission, the overall map hasn’t changed in any significant way since the sixties.

But unraveling the tangle of flight paths (if that were even possible) would not single-handedly solve New York’s airport problem. There’s still the issue of parking space. “Every pilot knows that it’s concrete, not airspace, that puts the final limit on capacity,” writes pilot J. Mac McClellan in a recent Flying magazine column. Right now, JFK has a combined nine miles of  “concrete,” including 13R, the second-longest commercial runway in North America. Theoretically, it can process up to 100 planes an hour in perfect weather; La Guardia can handle 75, Newark 108. That may sound impressive, but these are perfect-weather ratings. For instance, JFK’s “call rate,” an optimal realistic number of operations per hour, is 87, and the cruel reality is closer to 65. Vehement opposition from local residents and their representatives makes further ground expansion next to impossible.

One would think that the limitations of the system would telegraph something to the airlines—that it’s time to stop adding flights. No such luck. There are no mechanisms in place to regulate the number of planes carriers can schedule at JFK or Newark (La Guardia is capped at 62 to 64 planes an hour in bad weather). Neither is there an economic disincentive: Landing fees are cheap—$800 per flight for a large jet. So like all good capitalists, carriers take the silence of the law as an invitation to make as much money as possible. Passengers want options, and the airlines are in a veritable arms race to please them. This month, for instance, JetBlue offers fourteen daily flights to Fort Lauderdale (nine from JFK and five from La Guardia); Delta, trying to keep up, has nine. All told, there are an astounding 68 flights a day from New York’s three airports to Fort Lauderdale. One side effect of such generous scheduling is the shrinking plane size. On a recent Wednesday evening, a little after five o’clock, I was on the phone with JFK air-traffic controller Barrett Byrnes. “We got ten RJs [regional jets] on the Delta ramp right now,” Byrnes told me, when I asked him to describe the view from his position. “Ten 50-passenger jets. That’s 500-passenger capacity. You can replace them all with two 767s.”

For the sake of this discussion and our frayed nerves, let’s set aside the fact that cramming aircraft into a queue until planes are bobbing in each other’s wake is dangerous. Let’s focus on the absurdity of using perfect-world conditions as a yardstick for how many takeoffs and landings to schedule. This means that when something goes wrong—and something does, daily—each delay reverberates indefinitely. Should the weather go bad for just 30 minutes, causing, say, ten fewer planes to take off, controllers can’t slot them back into the system without pushing back ten other planes. And yet, airlines’ schedules are blithely calibrated to the best-case scenario. John Hansman likens this impulse to overschedule to the “tragedy of the commons” paradox: In a public area where everyone is allowed to graze his cow, overgrazing kills the grass. Everyone loses.

The situation is difficult, but is it unsolvable? That leads us to the final problem in the airports debacle: Flying is a jurisdictional nightmare, an impossibly convoluted tangle of private and public interests. Airlines are for-profit businesses doubling as vital necessities; generally speaking, they want as many flights as they can get. The Port Authority operates the airports on the most basic level—you want to put in a Cinnabon at La Guardia, you talk to them—and is responsible for runway upkeep and the like but has no say in airborne matters such as route design. The FAA reports to the Transportation Department, which makes all the big decisions regarding routes, fees, and caps and is itself hopelessly partitioned into dozens of divisions that meet only at the top. An observer may, in fact, get the impression that most of the FAA’s time is spent moderating between the airlines and the airports, and sparring with its own employees. “We land the planes,” offers the administration’s spokeswoman, Arlene Salac, when I ask her to describe what the FAA has clear, sovereign control over.

Since you can’t fly a kite in this atmosphere without getting into someone else’s jurisdiction, every major improvement initiative gets bogged down in interagency squabbling. The typical round of discussion goes something like this: A consumer group, a mass-media outlet, or Senator Chuck Schumer demands that the FAA do something. (“If you want to know why these delays occurred, three letters describe it: FAA,” the senator said in October.) The FAA then claims to be researching solutions under the DOT’s aegis, and proposes, say, redesigning some air routes. With every proposed solution, potentially affected constituencies start to complain, prompting the mass media, consumer groups, and Chuck Schumer to demand that the FAA do something. Replay sequence as needed.

To cut through the clutter and infighting, Schumer and others have been pushing for a permanent East Coast air czar: someone to whom individual FAA division bosses, airline reps, consumer groups, the Port Authority, and unions can answer. The controllers and the airline lobby like the idea, as does the Port Authority: “Someone with an elephant-thick hide needs to get in and break down these barriers—between the FAA and the controllers, for instance,” says May. But even without an elephantine hero in the wings, several solutions are being bandied about.

A long-overdue streamlining of the area’s departure and arrival paths could do wonders to ease delays. The Port Authority is proposing (aboveground, all it can do is “propose”) a second major air corridor parallel to the J80. Another plan, called “capping and tunneling,” seeks to put some flight paths directly under others, with clearly defined “floors” and “ceilings,” the way Teterboro’s lie under Newark’s. This September, the FAA finally pushed through a redesign plan that would reduce delays by 20 percent over the next five years. Among other things, it would improve final-approach courses for Newark and La Guardia, attempt to solve the conflict between JFK and La Guardia routes, and establish a new arrival path into Philadelphia. Sadly, however, all attempts at airspace redesign tend to run into the most predictable trap: nimby concerns. Already, there are five lawsuits filed against the FAA’s redesign plan, in Rockland County, Fairfield, Elizabeth, Bergen County, and suburban Philly. So far, there’s no injunction, but the combined power of the lawsuits, and the fact that outraged suburbanites make an irresistible voting demographic, could easily hamstring the plan.

For many years, FAA rules required that planes be kept at least three miles apart. Last year, buckling under the volume of aircraft in the air, the FAA changed its guidelines to allow 2.5 miles of separation. NextGen, a space-age system designed to replace the current network of radar towers with GPS satellites, is supposed to allow planes to fly even closer together, which would allow more aircraft in the sky at the same time. The system, which comes with a price tag of $15 billion, was originally supposed to arrive at the nation’s airports by 2015. These days, 2025 seems a likelier time frame for its completion. Dean Iacopelli, the president of the air-traffic-controllers union, doesn’t buy it: “From the perspective of an air-traffic controller working in New York, no technology exists that could let me handle more planes in this airspace.” JFK controller Byrnes is skeptical for different reasons. “It’s gonna be a heck of a reliable system,” he admits. “But if you have GPS in your car, does it make you get to work any faster? The traffic slows down, you slow down.”

The Port Authority has the power to build new runways, but the immutable fact is that the area’s airports may have reached their natural limits. JFK has been expanded over the marshlands seven times. To make it any bigger, says an FAA official who wishes to remain anonymous, “we’d have to condemn a bunch of buildings in the Rockaways.” That, needless to say, is a project even Bruce Ratner couldn’t ram through without causing some kind of uprising. All the agency can offer at our three airports in the near term is more “holding pads”—the idea being that idling planes might as well get serviced while they wait.

Some say salvation lies 65 miles north of Manhattan, near exit 17 off the State Thruway. There, the Port Authority has just triumphantly taken control of Stewart International Airport, an underused New Windsor airfield, and is planning to give it a ten-year, $150 million makeover. Officials are positively giddy about it: “There are two runways immediately available!” Even if the project to build a high-speed rail between Stewart and Manhattan is, for now, a pipe dream, the fourth international airport can still take some of the pressure off the system. Eleven million of the area’s customers live upstate or in New Jersey’s Bergen County—closer to Stewart than to any of the Big Three. JetBlue has already begun offering flights out of the airport. ATA, the airline lobby, is less impressed: “God bless them,” says Jim May. “But let’s not forget it’s managed by the same tracon [control center] as the rest of the local airspace. They still have to go south to negotiate airspace.”

Theoretically, one could persuade the airlines to use bigger planes. In fact, this was one of former FAA head Blakey’s pet ideas, and some carriers have been enthusiastic about it. A shift to larger aircraft, however, isn’t a panacea. Aside from superstar cities like New York, Paris, and London, the only way to make most transatlantic destinations economically viable to fly to is to make them easily reachable from smaller communities; thus, the hub-and-spoke system, wherein one takes a light jet from, say, Wilkes-Barre/Scranton to JFK and a 747 from JFK to Frankfurt. As jumbos require stronger runways than those found in smaller airports, and a direct Scranton-Frankfurt flight isn’t viable, a switch to bigger planes may leave our Scrantonite out of the loop.

There’s another drawback to larger planes: bigger wake turbulence. “Since Blakey dropped the ‘schedule less flights’ bomb, Continental Airlines has started to consider bigger planes as a means of decreasing the amount of flights,” reads a recent post on FAA Follies, a message board for disgruntled air-traffic controllers. “Wake-turbulence roulette at its finest.” Indeed, when American Airlines Flight 587 crashed on takeoff from JFK almost exactly six years ago, a major factor was a vortex created by a Boeing 747 in front of it. Because so-called wingtip vortices from large planes can hang in the air for almost two minutes, larger jets would need more space between them, which partly defeats the purpose.

This is the latest, and most controversial, weapon in the federal arsenal. On October 19—two weeks after its chief swore in a USA Today op-ed that “caps aren’t the answer”—the FAA announced it was going to ask airlines to limit flights in and out of JFK to 80 per hour for most of the day, and 81 in the evenings. It would mean that JFK, which has twice as many runways as La Guardia, would be handling fewer planes than its little Queens cousin. In fact, the cap would reduce activity at the airport to levels not seen since 1969.

The Port Authority has reacted with barely contained fury. “This is an economic-policy disaster,” says executive director Anthony Shorris. “It defies anyone’s sense.” He calculates that the FAA’s cap plan would cost the region 7,000 jobs and half a billion dollars in lost economic activity. The Port Authority is not entirely impartial—it stands to lose up to $200 million in passenger fees should the plan pass—but it’s not alone in its objection. Governors Jon Corzine and Eliot Spitzer co-authored an angry letter to the DOT, calling the caps a potential “crippling blow” to the region. The emerging common wisdom is that the plan would needlessly strangle the region’s biggest airport and spill flights into Newark (where, you’ll recall, the number of flights isn’t regulated), and they’ll still clog the same general airspace. Newark could eventually get capped as well, but even that won’t help, according to some. The number of flights in and out of La Guardia is already restricted. “And it’s the second-most-delayed airport in America,” points out Shorris. “Which should give you a hint that caps don’t work.”

Airlines aren’t wild about caps, either (Continental, with its Newark hub, is especially freaked out). In addition to reducing the total number of flights, there’s the question of how to distribute the remaining ones. So-called historical distribution of slots to established JFK clients would close the market to emergent players. There might never be another JetBlue, which, for all the side effects, has revitalized JFK and spurred competition in terms of both price and service. Another proposal—allowing airlines to buy, sell, and auction off their slots from a finite total number—would likely lead to excellent service to the parts of the country with the most lucrative airports and miserable service to the rest. Who’d want to waste a valuable slot on a marginally profitable flight to Lansing, Michigan, when you can add another one to highly profitable Miami? The cost of flights to many cities would likely skyrocket.

It’s anyone’s guess as to whether caps—or any proposed solution—will be implemented now, later, or never. “But the very fact they’re focused on the issue is progress,” says Port Authority chairman Anthony Coscia. That, in the airport world, is what may have to pass for optimism. In the meantime, your only option is to game the flawed system, which is what the next nine pages are all about.

Taxiing at La Guardia on October 29 at 12:02 p.m.
Scheduled departures provided by airlines; actual departures provided by

1. Air Canada 745 to Montreal
Scheduled: 11:10
Actual: 12:32

2. Colgan Air 4806 to Syracuse
Scheduled: 11:50
Actual: 12:20

3. US Airlines 2173 to D.C.
Scheduled: 12:00
Actual: 12:27

4. Midwest Airlines YX4 to Milwaukee
Scheduled: 11:20
Actual: 12:18

5. Mesa Airlines 7209 to D.C.
Scheduled 9:50
Actual: 12:22

6. American Eagle 839 to Toronto
Scheduled: 11:40
Actual: 12:15

7. Comair 606 to Savannah
Scheduled: 11:45
Actual: 12:14

8. Piedmont Airlines 4251 to Buffalo
Scheduled: 11:26
Actual: 12:12

9. Northwest 177 to Memphis
Scheduled: 11:10
Actual: 12:09

10. Delta 1918 to Boston
Scheduled: 11:30
Actual: 12:08

11. American Airlines 321 to Chicago
Scheduled: 11:35
Actual: 12:06

12. Comair 462 to Columbus
Scheduled: 10:55
Actual: 12:05

Additional reporting by Emma Rosenblum.

Gridlock at 30,000 Feet