Wednesday, November 01, 2006
How the Airlines Have Screwed-Up the Hub-and-Spoke System (GEEK POST)
[Disclaimer: On the Walkman Geek Scale, this post is definitely a 10]
In today's airline climate, the way that many of the the "Majors" (American, United, Delta, Northwest, Continental, and US Airways) employ the hub-and-spoke system is, quite frankly, retarded. It's why most of the Majors are in DEEP financial trouble right now.
How Hubs Work
By way of background, the hub system has been the centerpiece of the airline industry in this country for a long time now. The idea behind it is simple: At roughly the same time, many flights arrive at the hub airport from all over the place. The connecting passengers get off their flights, they disperse throughout the hub, and they get onto another flight, heading to their ultimate destination. All those departing flights then leave the hub at roughly the same time.
The benefit of the hub system is that it allows people to travel from virtually any city in the U.S. to any city in the U.S., with only one stop. If the airlines tried to connect every city to every other city directly, they'd be out of business. Consider the following example: How many people do you think would buy tickets for a non-stop flight from Burlington, Vermont to San Antonio, Texas every day? One? Perhaps two or three at the most? If you waited to sell enough tickets to fill a whole flight (even on one the smallest regional jets), you'd only be able to get from Burlington to San Antonio once a month. Not very convenient.
In contrast, if everybody who's flying out of Burlington (regardless of where they're going) takes one Delta flight to Delta's hub in Atlanta. In Atlanta, the two or three people who want to go to San Antonio board the flight to San Antonio. The five people who are going to Los Angeles get on the Los Angeles flight. And so on. In the end, it allows people to go from almost anywhere to almost anywhere.
What's Wrong with Hubs?
Airlines Needed to Create the Shortest Itineraries to be Competitive.
Historically, the centralized computer systems that listed virtually all airline flights (these are the systems used by travel agents and some of the online flight search engines), listed these connecting flights in order starting with the shortest total travel time. Assuming price is the same, people chose the first (and shortest) set of flights on the list. This created an incentive for the airlines to reduce the total travel times for as many of their itineraries as they could. They'd do this by scheduling as many inbound flights as they could to arrive at one time and they'd try to turn those planes back around to depart for their destinations as quickly as possible. Each one of these rushes is called a "departure bank" in industry parlance.
So far, this seems fine, right?
The Costs Associated with Turning so Many Flights at the Same Time are Immense.
Let's say I told you that you had to run a McDowell's restaurant for lunch one day. During the course of the three-hour lunch, you'll need to serve 50 meals (if hydrogenated imitation beef product and sawdust served on a bun-shaped slab of Styrofoam constitutes a "meal"). Let's also say that you know in advance that the 50 customers are going to trickle into the restaurant steadily throughout the three hour lunch.
Considering all these facts, how many McExploited-Employees would you need to employ to run the McDowell's for lunch that day? How many registers would need to be open? How much cooking equipment would you need? All in, how expensive would it be? The answer is simple: Because of the the stream of customers would be rather slow throughout the lunch, you'd need very few employees; you'd probably only need one register open; you'd need a bare minimum of cooking equipment; and in the end, the whole operation would be quite cheap.
Let's change the hypothetical a little bit. Instead of the 50 diners showing up in a steady stream during the three-hour lunch, let's consider what would happen if they all showed up within the same 15-minute period. To serve all these people without making them wait forever, how many employees would you need? How many registers would have to be open? How much cooking gear would you need? And compared to the steady stream of customers arriving, how does this compare in price? This answer is also pretty simple: You'd need scores of employees (perhaps chained to their workstations, per McDowell's company policy) to prepare all the meals within a couple of minutes; you'd need a dozen or more cash registers open; and you'd need a huge amount of cooking equipment to get the "food" prepared. In total, this scenario would be far, far more expensive for you to operate than the "slow trickle" scenario discussed above, even though the revenue you'd generate under both scenarios would be exactly the same. You'd make way more money with the "slow trickle" than with the "everybody comes at once" scenario.
As addressed above, the hub system is designed so that during each departure bank, all the planes land, unload, re-load, and depart at roughly the same exact time. Sound familiar? This is identical to the "everyone comes at once" scenario above. The airlines need more gates to fit all the planes. There needs to be a baggage crew to unload and load every plane, a cleaning crew for each plane, a refueling crew for each plane, gate agents at every gate, a catering crew (also called "provisioning") for every damn plane, a tug for each plane, and baggage equipment at every gate. All of this costs money--buckets and buckets of money. What's more, all of the expensive infrastructure and human labor sits largely idle (for as much as 30% of the day) in between the departure banks. As a result, the cost of maintaining a traditional hub with tight departure banks is incredibly costly.
In contrast, many Low Cost Carriers (called LCCs, including Southwest) have neither hubs nor departure banks. Rather, they stagger the arrivals and departures throughout the day (just like the "slow trickle" McDowell's example). For this reason, the LCCs need far fewer gates, tugs, crews, gate agents, refueling trucks, cleaning crews, and luggage crews and equipment to operate the same number of flights. Thus the LCCs save huge money by not having hubs or departure banks.
Some of the Majors have realized the sky-high costs associated with hubs and departure banks. In response, a handful of airlines have eliminated departure banks and scheduled flights to arrive and depart on an ongoing basis throughout the day (this process if called "de-peaking" a hub). American de-peaked its Dallas hub; United (somewhat) de-peaked its Chicago hub; Continental (somewhat) de-peaked its Houston hub, and Alaska de-peaked its Seattle hub.
Without departure banks, the cost benefit to the airlines is significant. But that cost is essentially shifted from the airline to its customers; this is because a passenger connecting at a de-peaked hub is more likely to arrive at the hub and have a much longer layover than the tight connections associated with traditional departure banks. Good for the airline, but bad for the connecting customer who's in a hurry to pick up his mail-order bride before she gets seized by customs.
Without hubs, there are far fewer cities that are connected to each other with only one stop. Anyone who's ever flown Southwest on connecting flights can attest to this. Leave Cleveland with a stop-over in Nashville. Wait two hours and change planes. Fly to Indianapolis for a stop that didn't show up on your itinerary and remain on the same plane. Arrive at New Orleans two hours later.
At best, non-hub airlines (in Southwest's case, it's what's called a "point-to-point" carrier) can connect people to some destinations after a bunch of stops. But at worst, a lot of smaller markets are simply not serviced by LCCs and other non-hub carriers because it's simply not economically feasible to connect them to other cities.
Conclusion
Hubs aren't necessarily a bad thing; if not for them, most smaller cities wouldn't have regular air service. But the way that many of the Majors have used the hub system, particularly when used with departure banks, is stupid and financially draining. While there will always be a place for hubs, if travelers want cheaper tickets, they'll have to be prepared to spend more time waiting at airports between flights if airlines de-peak in order to shave costs.
