The problem isn't overbooking. The problem is not auctioning the seats higher when overbooked.
I've seen a number of overbooked southwest flights and people can't line up fast enough for a $300 flight credit + next available flight. I can't imagine it ever going over $1k.
IMHO if any airline should continue overbooking it's SWA. No first class, no assigned seats. Most of the flights are short range with multiple flights options every day. You're much more likely to agree to switch to the next flight if it means only waiting 2hrs.
Yea, I agree the problem is not auctioning seats high enough. But the proximate issue is that United didn't offer cash, but $800 in United Fun Bux that are worthless for most people. The vouchers they were offering were only valid at dates United decided. The vouchers expired after a year. And they are offered in coupons for which you can only use one coupon per flight, and the excess value is lost.
Practically all news outlets are reporting that "United offered $800 in compensation", but the reality is that it's closer to $0 in compensation. It's disingenuous to call it "compensation".
A better move on the airline's part would be to offer some amount of airline credit, and then half that in cold, hard cash, at least up to $2,700 credit / $1,350 cash, which is what they have to give you for an involuntary bump.
In the European Union, airlines are forbidden from offering vouchers as compensation for overbooking. They must pay cash.
Simple solution that fixed a consumer-hostile approach adopted by a cartel of companies. Sadly, the EU rarely gets credit for the great number of consumer protections it has established; instead people like to rant about egregious regulations that either don't exist or have been grossly misrepresented.
Yes, the voucher thing is nonsense. I've been on overbooked flights before where I could have easily taken the next flight instead. I never volunteer, because the vouchers are worthless to me. Offer me cash and I'll often be first in line.
The only time I've volunteered to be bumped was when BA offered $700 in cash (technically I got a prepaid VISA card) and a free hotel+meals. That may be due to the EU though (there are passenger rights regulations about compensation for delays caused by the airline)
Back in the days when I traveled a lot, I would ask for some sort of on-the-spot compensation - usually, put me in first/business class on the rebooked flight. Sometimes it worked, sometimes not.
You can ask them to pay you in cash as long as you don't accept the offer before hand. They're legally required to compensate you with cash if you ask.
Where "suckers" are people that don't like what you like?
I like to travel so I would rather take the $1000 voucher than wait for a cash offer that someone like you would jump on before I could get a chance.
I've taken in over $5,000 of free air travel over the years and always used it to visit family and see places I wouldn't have otherwise. In the meantime people thinking like you got nothing because it was a "sucker's" deal.
I suggest you look into the study of economics and pricing to understand how people can get different levels of value from things. The you might understand why auction systems like that aren't magically lining your pockets with gold.
I don't understand why there even is a limit. What happens if nobody takes the 10k? Someone gets beaten up again? If they are certain someone will leave for 10k why have the limit at all?
For $10k, you could probably charter a private plane. So unless every single passenger on the plane is a doctor flying out for an emergency brain transplant, there'd be no rational reason to decline it. For $100k, you could probably buy a small plane.
If they leave the plane docked until somebody agrees, then you could end up in a situation where you get to your destination slower if you stubbornly refuse. Then, you could argue that it's in their best interest to (beat them up if they don't move).
My objection wasn't even over booking. My objection was eviction after the passenger already sat down. To me, that was the unredeemable act. I'm also ashamed that the police department apparently did not accept responsibility. We often talk about how it is a systemic problem when we accidentally run rm -rf in production and find out none of our backups work but we turn around and say "Oh surely the police department is full of good hard working people if it weren't for a couple of bad apples" and completely throw some people under the bus. No, the organization has life of its own and things won't change if heads don't roll at the top.
Actually no, you can't, because they limit how much you can apply to a single flight. To use up $10k in vouchers, it's likely you'd need to make 20-100 flights, before they expired in a year.
I've received large (not $10k) vouchers from major airlines before, and the only restriction is the expiration date.
They are effectively the same as cash for that airline, and I have yet to see a VDB voucher that is significantly restricted.
Airlines do give away a lot of "good will" certificates during irrops (e.g. weather events) to upset customers, which function more like coupons than traditional vouchers.
I'm happy to be proven wrong, but checking with a few friends who between us fly all the major US carriers - VDB vouchers are effectively airline cash with an expiration date.
Like the other reply you got, I've received a multi-thousand (but not $10k) voucher from United before and the only restriction was use within a year. I also didn't have to use on a single flight, but could use the full value on one flight if I wished (which I did, flew paid business, but was upgraded to first to APAC).
A lot more easily when you realize what a Y-class economy fare costs (Boston to Hawaii next week? $2700+)...
Vouchers are typically only valid for full fare unrestricted economy fares or better, not V-class discount fares (display details when you browse your next flight).
Calling airline vouchers 'worthless to most people' is pretty stupid. The vast majority of the population would love to travel more and is constrained mainly by their budget. The proof of this is how Delta never seems to have problems and they constantly offer Delta dollars.
And then you have guys like me, who fly very rarely and only if no other options are available. Any voucher tied to flying is inherently worthless, because not only I've lost a lot of time but also the whole point of flying - being able to move quickly from one place to another - is defeated by the fact that I'm now stuck in the airport.
To top that off, if this happened during a holiday it means I'm getting a bad deal anyway - if I'd want money I wouldn't take vacation days in the first place.
Vast majority of the population wants to travel more but, of all things, the cost of a plane ticket is holding them back? No, this is not true.
I travel for pleasure about once a year. I can travel more if I wanted to, finances don't hold me back, but once a year is enough for me. Traveling for vacation is usually pretty fun but at the same time just stressful enough that I wouldn't want to do it more than once a year. Airplane travel is really inconvenient.
Plane tickets are just a small part of the cost of a vacation so a free ticket isn't magically going to make them afford extra travel. United isn't paying them for their hotel, rental car, and activities they are going to do when they go on this extra vacation.
Even if you do make use of the vouchers, you better hope that United flies to that place you want to go and you have enough time off saved up at work and your schedule works out to use that voucher before it expires. I know you get a year but people have schedules and usually only have windows where it works out to take vacation.
If you are a business traveler you better hope you actually like traveling for pleasure.
You are very disconnected from a huge chunk of the US population. Traveling with a family of 4 to visit extended family on a popular period like Christmas or Thanksgiving (i.e. when typical middle class people get time off) will cost upwards of $2000 for a round trip flight a thousand miles away or so.
Most of the US population doesn't get very much vacation time from their employer. The problem isn't the cost of airfare (which is a tiny portion of the cost of any vacation compared to hotel, meals, rental car, etc.), the problem is a lack of time.
That's wrong. Most people get 2 days off every week. If you have a good priced airline voucher that means you can make a weekend trip to NYC, have dinner, see a Broadway show, spend the night and then fly back.
Flying out a Friday night and returning Sunday is the most expensive component of a weekend trip like that.
Oh please. Traveling is tiring for most people, especially with TSA and the hectic mess that is airports, and the severe discomfort of modern airplace seating; I don't know many people who really want to fly somewhere on a Friday evening right after work (meaning you get to your hotel somewhere around midnight if you're lucky, and you have to skip dinner) and return on a Sunday (meaning you get home late, leaving you exhausted for work on Monday).
> The problem isn't overbooking. The problem is not auctioning the seats higher when overbooked.
This. Oh so very much this. United could have avoided this whole debacle simply by not imposing price controls on the market for seats when it became a buyer rather than a seller.
Could they have averted disaster with money? It's astounding that there were no volunteers when they offered USD 800 for a seat, 800 dollars is a nice sum. Every single passenger on that airplane thought there was a catch involved, they were getting again 800 dollars in airline vouchers, which everyone knows have big gotchas attached. The airline has a trust problem, which won't improve anytime soon.
My understanding is that the next flight was something like the following afternoon. Business travelers are basically not going to voluntarily take a later flight at all. With respect to vouchers, most people don't travel all that much so a 1 yr. voucher isn't worth much. But offer cash (or a voucher at 2x) and you'll get people at some point.
> Business travelers are basically not going to voluntarily take a later flight at all.
This. As a business traveler, I don't know how I would justify this to my finance department and/or management. It would take too much effort to justify the change in travel plans for a shitty voucher that I'd probably never be able to redeem.
I once received the $500 voucher, and as an infrequent traveller, it was completely useless. The voucher went unused and expired. I wish they had offered me cash.
Now, I am a frequent traveller and mostly travel on business. Now again, the vouchers are useless to me.
I think the airlines should offer real cash instead of vouchers. More people would be ready to take those instead of vouchers.
Regardless of how heinous it might be to someone that a whole flight turned down $800, I can guarantee there was a number for every single person on that flight that would have averted the disaster. Now some of those numbers were probably much, much higher than others but I guarantee everyone had a number they would have agreed to (barring extenuating family circumstances or the like).
I think there would be takers even for $800 if it was cash but other commenters have pointed out that united never offered money but rather travel vouchers in small denominations that you cannot use more that one of in a single transaction.
I can offer $100k in thanksgiving fun bucks as well every time I screw up majorly. It will be in denominations of $50 and you can use them towards future consulting with me. My consulting starts at $100 per hour with minimum of one hour. You cannot use more that one voucher per transaction.
It was $800 in vouchers for other flights, not cold hard cash I believe. And even if it was actually $800, that is not worth missing a day of work unplanned for many people- the Doctor they dragged off the plane who had patients to see thought so.
Well in this case the market is distorted by the fact that regulations allow the airline to not let people board for a much smaller cost. If the airline didn't have that legal out, they would be forced to negotiate terms with each passenger whose contract they wanted to break.
Without regulations the airlines will just put a "we reserve the right to uniterally terminate this contract with a full refund" clause in their contracts.
Status quo: offer increasingly higher flight vouchers until you hit the federally mandated rate of 400% ticket price cash rebate, capped to $1350.
Without regulations: offer increasingly higher flight vouchers until you hit 100% ticket price refund cap.
> For example, an involuntarily bumped traveler who arrives to his final destination more than two hours late (fours internationally) is entitled to an amount worth 400% of his one-way fare (capped at a maximum of $1,350).
There are no regulations forbidding United to offer more when a passenger is already seated, and that passenger thus can't be legally bumped (apart from because of safety reasons maybe).
I am well aware of how it works. What I said is that the incentives to offer more than what they are legally required to offer by the IDB regulations (which does have an upper limit) are significantly lessened since the IDB rules give them a "get out of contract free" card for that amount.
And they could have offered the passenger free United flights for life and it still would have cost less than this embarrassing and protracted PR catastrophe ever will.
PR disaster happens at 40,000 feet and gets recorded by smartphones. Since they're still in the air the footage hasn't been uploaded to the internet yet.
The on-board AI supercomputer makes the determination that the negativity of a plane crash would be less than if the footage leaks out so it decides to crash itself in order to maximize profits.
It isn't a simple auction. Not all passengers are equal. There are issues that make an airline not want to bump some passengers. For instance, on some international flights there are rules against unaccompanied baggage. So the passenger without checked luggage will be bumped in order to save time. Passengers travelling in groups are also preferable as they will more likely share a hotel room. That's why families get bumped so easily. And then there are connections. The passenger who will miss subsequent connections is an expensive person to bump. This seems to have been the case in the United fiasco. The man was travelling in a group (his wife was with him) and didn't have any checked luggage, else we would have heard about it being offloaded or returned to him in hospital, and had no connections to make. Such complications go beyond the simple cash payout, making a strait auction difficult.
Want never to be bumped? Travel alone but with a pet in the hold. They won't touch you. Or travel with someone in a wheelchair. They are also an expensive awkward person to bump. As I said to one gate person: "We cannot be bumped. My grandmother only has pills for the next 12 hours." That lit a fire under them. Grandma knew to play ball and suddenly looked very much older than she was.
Another solution is even more simpler. After all seats gets booked, they should clearly mark overbooked tickets as "probable" tickets and heavily discount it + refund it if not eventually available. Keep in mind that airlines don't make huge profits per flight. Giving away $10K to 4 passengers is like giving away 80 seats or about 10 to 20% of revenue which could be almost all of their profits for that flight.
Those airlines take a risk by overbooking, and that risk can be penalized by a net negative profit if need be. If you're allowed to gamble, you need to be able to lose.
Yup, and moreover like any good casino it's pretty clear that they did the math already. That's why you're getting $800 fake bucks: they crunch figures on how much of that $800 they'd be on the hook for. If they gave you cash, I bet the inflection point where this entire overbooking scheme is actually fiscally worth it, becomes a losing proposal.
This is how they do in Indian railways. We book tickets. If it's full, we get waiting list number which will be confirmed a day before or on traveling day. If not confirmed, it will be cancelled automatically and refund sent to your account.
The problem is overbooking, though. Corporations are getting preferential treatment by being allowed to sell services which they're not going to fulfill.
The boner people on HN have for objectivism really does get quite old.
> Corporations are getting preferential treatment by being allowed to sell services which they're not going to fulfill.
Does the contract of carriage say the service they're selling is a guaranteed flight? Spoiler: No.
I hate the air travel experience as much as the next guy, but for the life of me I can't understand the derangement that seems to take hold of people when they start talking about. All else held equal, reducing airline costs (for example, by increasing utilization as with overbooked flights) brings prices down. It reminds me of when everyone I knew started complaining about airlines charging for luggage. When I'd ask them "I never fly with luggage, why should I bear the cost (fuel, opportunity cost, etc) of you packing too much", the most I got was some sputtering about how they didn't see why prices would be affected, even from people who I know knew much better.
> The boner people on HN have for objectivism really does get quite old.
Don't confuse your economic illiteracy with other people's support of objectivism. The perspective you're describing isn't even particularly libertarian, let alone Objectivist (and no, I'm not personally a fan of either of those philosophies).
Contracts are irrelevant here because there is no competition in that market, you have to accept the contract or walk to your destination. For contracts to be meaningful the parties have to be of comparable power and both parties must have the opportunity to choose a competing good or service. This is plainly not the case here.
Sure? Except that has nothing to do with what I was talking about. The irrelevant statement here is yours.
I was specifically rebutting the claim that airlines are getting "special treatment" [by the government] by being allowed to sell services they don't fulfill. The service they're selling is not "a guaranteed flight".
Words have meaning and facts matter, beyond "Airlines bad, pick a side". It's possible to think that they're anti-competitive and shitty without making entirely false claims about them (as the gp comment did).
I'm not a frequent flye but I would really appreciate some clarification on this topic. How does an airline selling more than the number of physical seats it has increase efficiency/effectiveness more than if it sold the same number of tickets as the number of seats available? Do they have that big of a problem with people canceling flights? What are the return policies usually for flight tickets?
I've never heard of any airline overbooking here in Europe. Yet they seem to manage just fine. In fact Ryanair (a budget airline) has a bunch of flights advertised for around €15 ($16~ USD)...
it uses less fuel per passenger to ensure that every flight is jam packed. It's more efficient and everybody wins. There are almost no examples of situations that overbooking + auctions doesn't solve.
I think you don't understand the law as it currently exists, because what happened on that united flight was in clear violation of that law.
They didn't deny the man boarding, they deplaned him. This isn't something the airline is allowed to do for economic reasons, ever.
Can you explain what you mean by the overbooking + auction system? I don't fly often and though the news has left a bitter taste in my mouth about United, I do want to understand how their system works and their flaws in a less biased way.
One possibility is that the regional flights are only marginally profitable for United, and very slim margin for the regional carrier (Republic Airline/United Express) so they limit voucher amounts for bumped passengers.
For an Embraer 170 aircraft (70 seats) the operating cost of the flight is about $2500-$3000 for the Chicago to Louisville flight. A supplementary flight with the 4 crew and any re-booked passengers would have cost the airline about $5,000 if they flew out-and-back as a non-scheduled flight.
SWA 737s cost around $4,000/Hr to fly, but a non-scheduled (Out and back) flight on a shorter route would still cost less than $12,000 in operating expenses.
The $1000+ flight credits for bumped passengers add up fast, but supplementary flights are relatively rare for airlines, other than for cancellations due to maintenance issues. CEOs like Oscar Munoz need to make supplementary flights easier to authorize, and the negative PR would be significantly reduced.
They could have just used a rental car or a taxi. Airline crews don't have to travel by air. The amount of inflexibility displayed by United is breathtaking.
Needs to be a fast mode of transport to keep crew duty hours down, so ground vehicles are out. I was once on a plane at Dublin airport that had to wait for a few minutes behind a small jet that turned out to be the RyanAir (UK and Ireland low cost carrier) company LearJet which they use for positioning flight crews in an emergency or time sensitive situation. Surely United should have chartered a small jet to get crew there, should be doable pretty quickly?
Ryanair is an entirely Irish company, and is one of the largest airlines in Europe.
They are also fanatical about not being late or cancelling flights. I used to work for them, and there are posters everywhere about the 18 minute turnaround times (they recently expanded this to 28 mins, when they introduced pre-booked seats as standard).
As far as I know, the Ryanair Learjet is mostly used for spare parts. It's cheaper bringing a spare part via private jet than cancelling a flight with 180 passengers. For crew transportation, the 80 stations that Ryanair has in Europe should ensure that there's a spare crew only a few driving hours away
That's what I'm wondering about. Putting the 4 crew in an Uber or rental car would've probably cost $1k-1.5k. It's standard in Europe that crew gets long distance taxi rides throughout the country, why isn't that the case in the US?
Do you understand how expensive it is to take a taxi that far? Or the fact that taxi/rental car companies usually don't like you going over state lines?
Right, that's the ballpark of the 'natural ceiling' for how much they should be willing to pay (at maximum) for a solution involving deferring passengers; also in /real money/.
Rental car companies don't like you going over state lines? What? That's ridiculously untrue. I've never ever ever heard of anything like that and I never didn't take a rental car over state lines. I always read the fine print and I've used several companies.
It may not have been practical for the flight crew because of hour on duty restrictions. But they could have transported 4 passengers for $1-2 k on a limo.
I've occasionally had rental companies limit me to nearby states, but they usually don't care. When they do care, I'm certain they can be made to stop caring if you apply enough money to it.
Rental cars (booked in Germany or Switzerland) often included restrictions for me.
Most of the time they excluded a number of eastern/balkan states, quoting (if they offered reasons at all) insurance prices would be higher. The more expensive the car, the more likely were these limitations: Booking a Golf was 'go anywhere', booking a BMW 5-something .. not so much.
Completely irrelevant. We aren't talking about Germany or driving to other countries, that's totally different. We are talking about travel within one's own country. I would expect rental car companies to limit travel to foreign countries.
We = this article and the commenters on this article. We are commenting on travel between two US states, nothing to do with crossing country lines. The grandparent topic was, specifically, about ground transportion for four Americans from one part of the US to another part of the US.
Well, this whole subthread is about an American airline operating in the US. (And yes, you can potentially drive a car across an international border from the US!)
I've seen this happen many a time with British Airways in the UK too - they just keep offering more money until someone volunteers to move to a later flight. Seems like a good way to manage overbooking, and I've never seen it not work.
The difference is that in the UK, you'll either have multiple flights a day or can take train/car without much delay instead. That's very different in more sparsely populated countries as the US.
But cash is likely to be the main issue here. If united had offered $800 cash, they would've probably found someone to rent a car and drive down there instead of flying.
Often this is true, yes. But I've also seen this happen with flights within Europe on occasion too - the bidding usually starts with £100-£200 cash plus a night in a good hotel with transfers dinner paid for, and a flight the next day. I don't think I've ever heard the auction go very much past that. I guess there is always someone not in a hurry willing to extend their trip by 1 day. Actually, I think I considered it myself one time, but someone else took the deal while I was still thinking on it!
Compensation is definitely the main issue. If United had offered a decent sum of cash (I mean cash, absolutely not United vouchers) they would certainly find someone willing.
I need that car so my brother can drop his kids off at their grandparents house a state over. You don't want to give it back? Tough, the police are on my side. Thanks for your understanding. Here's an $800 voucher for the next time you buy a car off me.
"This case" was an unusual incident. The majority of overbooking is handled before anyone enters the plane.
So while your analogy works for the doctor who was dragged off the plane, it doesn't work for the overbooking system in the air travel industry as a whole.
Do people really miss their flights so often that overbooking isn't often an issue or do a lot of flights have people that need to wait for the next one?
As someone that's never flown it's surprising to me that overbooking is even a thing airlines do, who's out there missing their flights? I can't even imagine cancelling a pre-booked taxi never mind a flight!
Um... yes. I missed a flight due to major flooding on the route to the airport requiring me to take the loooooooooog way around. The flooding didn't effect the airport itself (or my flight) because I don't live in a town with an airport.
It's also very, very easy to miss a connector flight if the plane coming in is delayed, or even if the airline decides to route you through an entirely different city. I've missed a couple that way. You can get held up at customs or security when landing in a new country and that country requires everyone to go through customs and security again.
Then there is illness, accidents, and emergencies that come up all the time.
Yes. Mostly because of delays that cause them to miss a flight that they're connecting with.
You also have people oversleeping, getting caught in traffic, getting confused about their flight time. It happens. I've missed a few flights in my life for these reasons. (But just a few.)
Finally, business travelers do make last minute cancelations and changes on a pretty regular basis--for example, to get on an earlier flight out.
> Do people really miss their flights so often that overbooking isn't often an issue
Yes. Wendover Productions on YouTube posted an excellent video a while back explaining why airlines overbook their flights: https://www.youtube.com/watch?v=EqWksuyry5w Most of the time it's not an issue.
There can be a number of reasons for missing a flight. Two examples: a) You're going from an airport which you need 2h+ to get to, and your train is delayed. b) You're on a week long holiday going through multiple towns, you miscalculate travel time / miss a transport, and don't arrive on time.
The problem is overbooking. When it will be definitely banned by law, flying experience will get better. Want to pay less? No refund on no-show. Prefer flexibility? Pay something more. But, once you get top the airport, you know you'll be on the plane.
So... my first reaction to your comment was "why don't they just turn this into an actual auction, with no price limits? This seems like such a good solution that I feel like I must be missing something."
However, thinking about it more, I think there's a problem with that that's hidden by the nature of the United debacle.
In that situation, the passengers had tickets, and they company was trying to seat their own staff. That is, the potential buyer was the airline, and the sellers were the passengers. So in that scenario, auctioning makes sense, because you can literally flip the buyer and seller roles.
However, in a typical overbooking situation, there are more passengers than seats, so it's unclear who the buyer and sellers are, because none of the "sellers" actually "own" anything definitive. If the tickets are all infinitely valuable to each passenger, no one can get on the plane, and there's a sort of stalemate. Then I suppose there would be a cost to each passenger, and an incentive to accept a price, and it wouldn't be infinite anymore. But in general, the cost dynamics seem weird to me. But then again, I'm not an expert in this.
At some level, overbooking seems indistinguishable from fraud--which can also be profitable for its practitioners.
> However, in a typical overbooking situation, there are more passengers than seats, so it's unclear who the buyer and sellers are, because none of the "sellers" actually "own" anything definitive.
I'm not quite sure I'm following. When the airline overbooks then tries to pay off passengers to accept a different flight the buyer is always the airline.
They're buying your seat back from you to satisfy their contract with someone else. I can't fathom a situation where every seat is infinitely valuable - out of several hundred people surely at least one person will accept a million dollars for example.
Overbooking can look like fraud, but if the airline offers an uncapped ever-increasing payout for a seat then the risk profile increases significantly for them and they'll do it to a lesser degree.
In the U.S., plane tickets are a contract where all passengers agree to the standard overbooking process: passengers who don't have a seat are placed in a standby status. If they are bumped from the flight, they can get up to a certain amount of money and are placed on the next available flight.
It's shitty because no airline offers an alternative to this contract, so all passengers have to accept it... until Southwest decided to stop overbooking.
> However, in a typical overbooking situation, there are more passengers than seats, so it's unclear who the buyer and sellers are
The passengers with tickets are still sellers in your second scenario; to the airline as a buyer. The airline needs to buy back the number of seats that they oversold for which passengers checked in.
Southwest Airlines has the best website, best checked luggage policy, best change policy, and now no overbooking. The big four domestic airlines all treat coach passengers like cattle, but Southwest is the least predatory.
Edit: Also, consider that Southwest is very profitable despite the fact that they don't have first class or business class sections on their aircraft. Yes, passengers pay different fares. But as far as I know, they all sit in the same crappy seats.
To offer a counterpoint, Southwest also offers the worst boarding procedure, unless travelling alone or with kids. My wife and I were going on vacation and it was impossible to get seats together as we were pretty far back in the assigned line positions and by the time we got on the plane, there were no two vacant seats next to each other.
After landing we discovered that checking in at the right time might get us further ahead in the line but we tried that, checking in within minutes of checkin opening and were still pretty far back in the lines.
We ended up buying an "upgrade" to get us ahead in the line and next to each other. For us to do that for both flights, we'd have been paying ~50% extra for a feature that comes by default with all other airlines.
Also, while interacting with Southwest's famously "better", more "caring" staff, we were met with total disinterest and absolutely no sympathy.
> After landing we discovered that checking in at the right time might get us further ahead in the line but we tried that, checking in within minutes of checkin opening and were still pretty far back in the lines.
"Minutes" isn't good enough. I travel Southwest with my wife all the time and we have never failed to get seats next to each other. You need to go the the check-in form 24 hours and 2-3 minutes before your flight, have your confirmation number and other info filled out and click "Check-in" the exact second the clock turns over to 24 hours before your flight. That will probably get you a low-numbered B or a high-numbered A boarding group.
The reason that checking in earlier doesn't get you the first spot in line isn't that the tickets go really fast like concert tickets. It's because customers who are on a connecting flight can check in 24 hours before their first flight and therefore more than 24 hours before their connection, giving them priority. I almost always have a better seat on my second flight than my first on Southwest.
This isn't a bad strategy by them as it tends to give better seats to people on long itineraries who have been waiting around, and since those people fly long distances they are also high-value customers.
I agree 100%. I often hear people talk about how Southwest offers such a great experience. I find the boarding process frustrating (and no more efficient, unfortunately) and the staff no better than my usual airline—either far less friendly, or hokey in a very contrived way.
Especially on Delta and American, I don't even get brusque professionalism, I get outright rudeness. The longer I've lived in the US, the more I've learned to minimize conversations to flight attendants to "I'll just have water", "thanks" and "have a nice day".
AirTran (now part of Southwest) used to be really nice.
It's always interesting to me how much experiences vary, and how likely it is that our biases also color our perceptions. Almost always when I get on a Delta flight (admittedly I fly them the most), I'll get a friendly hello and what seems like a professional and genuine smile when I get my beverage. My last few experiences on Southwest have been a borderline sneer (if any acknowledgment at all) as I get on the plane, followed by bizarre jokes over the intercom or a round of applause for the plane landing (?).
> My last few experiences on Southwest have been a borderline sneer
I've had the complete opposite experience. I used to frequently fly southwest between Denver and PDX, and the flight crew were chill AF. You can almost guarantee a free drink or two on a southwest flight if you ask nicely -- I can't say the same about any other domestic American airline.
Counter-counter point: Southwest boards faster, by far, than airlines using the standard pre-assigned method. That means less time sitting on the plane, or waiting to board the place. Also means less abuse of selling seat upgrades.
I listened to a radio interview many years ago with a physicist who had done simulations to find the optimal boarding strategy. Once he explained it it was totally obvious. The biggest time waster is waiting for people in front of you to finish putting up their luggage so that they step out of the ail and you can get to your seat. The boarding strategy most airlines employ of boarding people at the same time who sit in the same part of the plane minimizes opportunities for passengers to put their luggage away in parallel. So it's actually a actively negative boarding strategy. Boarding in random order would be better. Ideally you would space people out from different sections if the airplane and within reach group someone from the back of the plane would board first. They'd all arrive at their seats in parallel and start putting their luggage up at the same time and then the next group repeats the procedure. I think what Southwest does is interesting because it leaves the process in essence to the passengers. I've seen it go both ways where someone takes the first free seat and starts putting up their luggage, blocking everyone else from making progress. Sometimes people are great and walk by available seats top the middle and now people can start putting their stuff away in parallel.
Ever since I listened to that I cannot stop thinking about it when boarding is slow and get upset about the airline making things actively inefficient. A friend made the good point though that a likely reason for the process being chosen despite actively wasting time is that it feels intuitively efficient to most passengers and keeps them happy.
Mythbusters also did a segment on boarding efficiency. Their conclusion that "back to front" boarding was the slowest, and "no assigned seats" was the fastest, but that passengers hated boarding with no assigned seats. Window-middle-aisle boarding was second-fastest.
My experience is that American passengers are just terrible. Planes here in Japan board so quickly in comparison, with a standard back-to-front boarding.
Is it accurate to say that Japanese are just used to orderly boarding in general? I haven't visited yet but just watching the amount of order and efficiency at train stations on YouTube is crazy compared to the US.
In my limited experience from boosting Japan I'd say that Japanese public life is generally much more organized than in the US.
At the train station there are markings indicating were the doors of the train are gonna be and people line up at the markings. It makes it really easy to pick a good spot to wait for the train and makes boarding so much less stressful. At least for me. In SF train stops I'd always be strategizing where to position myself and keep repositioning as more waiting passengers are bunching up in certain areas. I'm crazy that way. So Japan was very relaxing for me.
Southwest only sells tickets directly (another way they keep fares low) and they don't share flight data with Kayak/Priceline/etc..
If Southwest covers your route, their fares are almost always going to be competitive with the lowest price. Where Southwest is not good is when you have longer flights (cross country) which will almost always entail stops and or changing planes. But it is great for short stuff like flying between OAK/LAX/SFO/PHX/LAS.
The seat upgrade procedure is pretty nice. $20 for priority boarding on southwest vs. multiple times that for a "economy deluxe" seat on other airlines? Sign me up.
It also would seem to mean they can either schedule flights closer together or have additional buffer to prevent cascading flight delays, either of which is a win.
My proposal is that southwest allow people traveling in pairs to board before the C group, and require that they sit next to each other.
On my last flight, I was traveling with a four year old, but my first flight arrived late, so we were the last ones on the connection. There were only middle seats left (and the flight attendant was rude about it). A woman and her husband were sitting in A & C seats, and one of them moved for us. But they could have sat together if everyone had just sat more efficiently.
Selling the right to sit next to your travel companions seems much sleazier to me than selling other airplane perks, like more luggage or snacks or specific seats.
If you check in as soon as you can (within about 5 mins or so), you get either the end of A or the very beginning of B. I've always been able to get a row to myself as long as you check in early.
Also, early bird check in is $15. Only one of you would need it (no one will fight you for a center)... So $30 for a round trip if you don't want to have to check in on time.
Southwest's procedures made more sense when they only did at-gate checkin. You boarded the plane in the same order as arriving at the gate, and your boarding pass was just a reusable plastic card with a number on it.
They had to abandon this when the TSA started requiring a boarding pass to get through security, which made gate-side check in impossible.
Yeah, even when told it's a full flight, watch single passengers sit Aisle, Window, all the way down the plane, leaving groups behind them to be spread across middle seats over several rows.
Would be better just to file people in, next seat, next seat, with exceptions for groups to go to the next row. "First available seat or seat group that will seat your group".
i prefer southwest because there's little incentive to fight for position. They give you a complete ordering, so you can wait until the last minute to line up.
(also as others mentioned, it's faster which is great too)
If you care so much about an aisle seat, get early check in, it's $15.
I try to be the last one one the plane even with assigned seating: it gets really hot and unpleasant on the tarmac, and I'd rather spend less time there.
Their boarding procedure is selected because it's scientifically proven to be the fastest (out of the acceptable options). There are more exotic procedures that are even faster but people wont put up with them. Ideally you want back-to-front, outside-in, staggered sides, staggered rows to make it as parallel as possible.
Southwest was a case study in a business class I took in college. Our materials credited their success to their management style of encouraging enjoyable, fun behavior from their employees and giving them the freedom to "do the right thing".
It may be a stretch, but I imagine if United employees felt they had the freedom to circumvent policy to "do the right thing", the situation could've been avoided.
I'm curious, are there stats for this that are adjusted for the number of flights/total time of flights/total number of passenger-hours flown? I found one list[0] but it didn't quite give that:
These stats are usually better for new airlines (in Europe and US at least) because air travel has become so much safer. Accidents are so rare now that it's just luck if you had 0 or 1 accident in the last decade.
Virgin America[0] has a better website, in my opinion. Southwest has always been nice to me, but when they had their big outage in May 2016 they were quite rude to everyone.
Alaska is a pretty nice airline actually with a fair number of benefits that no one else seems to offer. I hope they grow viable in Southern California.
I didn't follow the doctor-getting-dragged-off debacle closely but I have a question about the basic mechanics of what went down:
When I've been boarding on flights, they do the overbooking announcement at the gate and get volunteers to take later flights there, before people board. That makes a lot more sense because, in the event that no one volunteers, the airline can simply refuse to let certain people on.
They don't have to take the seat away from the passenger, they can just not to give it to them. Certainly, psychologically, people are a hell of a lot more attached to something once it's been given to them. I've toyed with the idea of taking a later flight before, but I'd be much less likely to do that once I was all settled into my seat.
How is it that the doctor was already seated when they "volunteered" him to not fly? That seems like the main fuck-up to me.
It is indeed the main fuck up - despite knowing they were 4 seats short (because of their internal requirements - it wasn't overbooking) and had no takers on their initial offer ($400), they let everyone board. You are not supposed to do that, especially if the number of seats is that significant.
It's suspicious that they didn't even make an attempt at getting to the maximum offer ($1300) before forcing matters. To me it looks like they were under pressure to avoid a big delay, so they rushed through boarding and just went "fuck this" when nobody picked up the $800 offer.
The important detail that everyone likes to ignore is that the flight was not overbooked. It was filled to capacity. Everybody who wanted to be on the flight was on the airplane. Then United decided to reaccommodate four seated passengers to make room for four employees who were needed elsewhere.
Their contract of carriage allows them to kick you off under certain conditions, but that's not one of them. "Overbooking" didn't come into play at all in this case.
How is it that the doctor was already seated when they "volunteered" him to not fly? That seems like the main fuck-up to me.
It is. Different parts of FAA rules and even United's contract of carriage apply once you're boarded. Overbooking doesn't even apply here - the doctor had a reserved confirmed seat.
It's been demonstrated in countless other threads (1) and articles (2) how economically important it is to overbook.
Overbooking is central to airline profitability and their ability to provide flexibility - like partial refunds, or SW's "open tickets" policy on business select.
As a consumer I appreciate these things and find them incredibly useful.
United's response of "We'll push the allowance up to $10,000" is completely acceptable to me, that seems like the simplest solution.
I still don't believe this claim. A 727 seats 129, a 787 seats 290, and a 747 seats 366. That means it only makes flights slightly less expensive. I only travel a few times a year and I'd rather pay 2% more than get bumped off my flight.
Why do we allow airlines to overbook but not movie theaters or other industries? This seems like an artificial advantage we've given them and it doesn't seem beneficial at all to the average consumer.
Paying 2% more and getting bumped off your flight is a false dichotomy. United's approach of offering $10k avoids both.
I don't see a reason to waste seats if it can be avoided without hurting consumers. Wasting seats is inefficient economically and bad for the environment.
I think you're missing the fact that a percentage of people don't turn up for their flights, so on most overbooked flights, the airline doesn't have to buy back tickets at all.
If theaters had an escalating buyback price for their tickets, I could be persuaded to, say, not see the next Star Wars movie for $100 and then see it later. I have a price. Most people do.
Almost nobody gets bumped involuntarily on competent airlines. They just increase the bid until someone has a hell of a great day by clearing first class round trip tickets across the globe for volunteering to spend the night.
This isn't your point, but, in case anyone's wondering, Southwest doesn't fly any of those planes. An interesting thing about their operation is that they exclusively fly Boeing 737 aircraft.
Overbooking is a technique used by legacy airlines. Pure LCCs don't use it (certainly in Europe) because their tickets are non-refundable other than from EC261 claims and they don't have inter-lining or partnerships with other carriers. I'm surprised Southwest were overbooking in the first place given they're a budget airline...
Edit: An exception may be Easyjet (though I've never seen it happen and I fly them a lot). They offer flex tickets mainly aimed at business flyers which would make it worthwhile for them to do some limited overbooking.
Economically speaking, non-refundability doesn't necessarily make overbooking unnecessary.
Let's take two airlines, A and B. Both only sell non-refundable tickes. A never overbooks. B always overbooks. What happens?
A will fly with some number of empty seats on average, because they still get the occasional no-show. B flies with fewer empty seats on average because they overbook.
Which airline can offer more competitive prices? It's B. Because by overbooking B's revenue per plane leg is higher.
If that isn't clear, here's an example. Say the planes have a capacity of 100, and on average 2 people fail to show. A sells 100 tickets for $100.00 each. B sells 101 tickets for $99.02 each. On average, B makes more money on every flight, _and_ is more competitive on price.
Of course with the numbers in my example the compensation paid by B on the odd occasion that everyone shows up probably kills any additional profit. The real numbers matter in order to determine to what extent this really happens in practice. But I hope I've demonstrated that non-refundability of tickets doesn't necessarily or automatically mean that overbooking is not profitable.
So there is some potential level of competitive benefit to overbooking even on non-refundable tickets, assuming that some people still occasionally fail to show (because they miss a connection, become ill, or whatever).
On the other hand, the discussion delays boarding, you need better/more gate staff, you need the infrastructure to pay out, quickly book to other flights, etc.
I think Ryanair's approach of not overbooking is sensible as it keeps the operation very easy. Also, they can book every flight to 100%. If you deny boarding, you need to offer alternatives. That's not possible if you're the only airline flying from the airport (Ryanair used to use a lot of regional airports) and your next flights are already at 100%.
> If that isn't clear, here's an example. Say the planes have a capacity of 100, and on average 2 people fail to show. A sells 100 tickets for $100.00 each. B sells 101 tickets for $99.02 each.
As usual, people striving to save one dollar no matter what (in this case, giving up the certainty of flying) are the problem.
It could be much more than one dollar. Those were my made up numbers to demonstrate the principle, not any kind of indication of the magnitude of the real numbers.
> As usual, people striving to save one dollar no matter what (in this case, giving up the certainty of flying) are the problem.
It's not necessarily a problem. Flyers can choose between certainty of flying and other nice things that cost money. Most appear to choose having other nice things instead. Southwest is banking on flyers choosing the other way. There may well be space for both approaches in the market.
Not explicitly, but I'm guessing that if you've paid for business/first, there is some small chance you'll be bumped back to economy (as in the story making the rounds a few weeks back) but probably a vanishingly small chance you'll get bumped off the plane entirely.
Southwest flights aren't that cheap unless you book in advance. And they used to (still might?) have a very flexible refund policy. You could not show up for a flight, never call, and still get a full credit for the value of your ticket to any other Southwest flight in the next 12 months w/ no penalty. I really like Southwest.
They've restricted it slightly: now you have to notify them at least 10 minutes before the flight. But still no penalties - I really hope they don't have to do away with that policy since that's the main thing I like about Southwest.
The 10 minutes prior departure cancellation rule is for full refund to the original payment method, i.e. refund to credit card or cash. If you missed the flight without cancellation, you got back SW credit as refund, which has to be used within a year.
For the lowest-price ticket, "Super Web Saver" or whatever, there is no way to get a cash refund for a cancelled flight 24 hours after purchase. You just get the credit for cancelling 10 minutes before departure.
After that, you get nothing. It's in the contract. I think it's only been that way for 2 years or so.
They recently added quite a few international flights [1]. They now offer service to Aruba, Belize, Cabo San Lucas/Los Cabos, Cancun, Cuba (not sure if it has started yet), Grand Cayman Island, Liberia, Costa Rica, Mexico City, Montego Bay, Nassau, Puerto Vallarta, Punta Cana, San Jose, and Costa Rica.
The majority of seats sold in Europe are now one-way / singles, due to the rise of low-cost carriers which don't offer anything else.
However it's still a good idea to book each individual flight in a different transaction. Avoids any shennanigans on the airline's part for no-shows and there's no downside for the passenger other than a slightly longer booking process.
The article pretty clearly suggests this policy was under consideration anyway, though the United incident at least seems to have affected the timing, and might have made it easier to pull the trigger.
> Overbooking is central to airline profitability
Flying full is certainly beneficial to profitability, and overbooking makes that more likely. But it's not without costs, and it's quite possible that the cost/benefit has changed since the policy became a norm.
The main budget airline in Europe (Ryanair) doesn't overbook flights, and as far as I am aware never has done. They managed a net profit of 1.4 billion dollars last year.
Yep, the main difference in Ryanair's model is that basically all tickets are nonrefundable. This way, even if the plane leaves half-empty on the day despite being fully booked, it will have paid for itself - which is what should happen, really.
What you lose by renouncing overbooking is the ability to change your ticket without paying for it again (which is what happen with Ryanair - fees to alter bookings are so huge, it's cheaper to just buy a new ticket).
That's not really an argument. Profits could have been even higher if they had leveraged overbooking. And some of Ryanair's tickets are certainly below marginal costs - making me wonder if really all fights pay for themselves that easily.
My guess why they don't use overbooking: it's really complicated. Both finding the right balance of overbooking as well as managing the procedures at the airport. Ryanair has simply decided in favor of simplicity forgoing some additional revenue
> Profits could have been even higher if they had leveraged overbooking.
But risk would also be much higher - as you say, handling overbooking correctly is complicated, and when it goes wrong it can go very wrong. Ryanair didn't just renounce revenue, they renounced risk and the costs associated with managing that risk.
> And some of Ryanair's tickets are certainly below marginal costs
The operational word being some. Much-fanfared rock-bottom prices nowadays are very limited, and the cost of a seat goes up very very quickly after they're exhausted. The cheap tickets are now a marketing proposition for the real ones.
Even their normal ticket prices are regularly cheaper than their competitors. I fly a short hop once every six weeks or so (edinburgh to Dublin - maybe 40 minutes). Even the full price tickets for the flight are almost always cheaper than their only competitor on the route (aer lingus). I've noticed this on other European flights too, where they are often 40-50 euro cheaper than their competitor.
Ryanair are running at >90% seat capacity across their routes, without overbooking. Not sure the complexity of introducing it would be worth the gain given their load factors.
Edit: then again, with their relatively recent business plus fare (which allows ticket holders some flexibility), who knows...
Oct - Dec 2016 Passengers Denied Boarding
by U.S. Airlines. United's Denied Boarding is 0.40 per 10k passengers vs JetBlue DB of 1.19 per 10k passengers.
JetBlue doesn't overbook, but they do swap A320 for A321 which have 50 fewer seats, fairly randomly, which leads to a lot of IDBs (you can ask for volunteers and get them when it's 1-5 people on a 200 person flight; you are probably screwed if you need 50 people on a 250 person flight.)
> Overbooking is central to airline profitability and their ability to provide flexibility - like partial refunds, or SW's "open tickets" policy on business select.
Are you sure of that?
I would think that airline profitability is largely dictated by competitive forces, rather than the particulars of in which practices the overall industry is permitted to engage.
(Note that I'm assuming a world where all airlines, or no airlines, overbook. If only some of them overbook, then I agree the practice could affect relative profitability of those who do / don't overbook.)
You're right that it's not intrinsically essential.
To an airline, a flight that leaves with an empty seat is basically throwing money away; that empty seat doesn't mean you can take less fuel, fewer flight crew members, or wear out the airplane less. Therefore, the empty seat saves the airline no money.
Furthermore, people often miss their flights.
Overbooking is one way to deal with these two assumptions. Assume a certain number of people will miss the flight, and sell some seats twice. Then you don't fly with money-eating empty seats, and you can lower the ticket price a bit, because you don't have to charge customers the cost of those statistically-likely empty seats.
There are other solutions. Don't offer any refunds or reaccommodation to no-shows. The seat was empty, but someone paid for it, so who cares? But customers probably don't like that. The might even like it less than overbooking. I don't know, I don't run an airline. There are options, but overbooking exists for a reason.
Well, there is more fuel left in the tank, so there is less on the refill so yes it does save fuel, not much but it does save some. I prefer no refunds for no-show policy myself, that is how it works when I buy a ticket to see a play. Now refundable upto 48 hours before the flight might be even fairer.
The fuel increase is marginal. You still have to burn fuel to carry that fuel to the next airport. And you have to carry the airframe with capacity for that empty seat. The 100 kg for a passenger+luggage is much less than the fractional weight of the plane required to carry them!
For some numbers, a 777-200 carries 313 passengers, weighs 135k kg, carries up to 100k kg of fuel, and has a max takeoff weight of about 250k kg. 313 passengers and luggage, averaging 100kg, weigh just 31.3k kg. Your ticket therefore pays for your weight, plus 5x your weight in airplane structure, plus (up to) 3x your weight in fuel.
If you don't show up, they're still carrying 8x your weight anyways, so saving 200kg isn't that important.
(PS: apologies for the 'thousands of kg' units. I thought it was more readable than Mg or fully written out numbers.)
> I prefer no refunds for no-show policy myself, that is how it works when I buy a ticket to see a play.
What about if you are flying long haul and linking up to a domestic flight on another airline? "Oh, we can't help you because your previous flight was delayed due to weather/mechanical/congestion.you should have flown on one carrier even if the price was 2x"?
Airport congestion is a big issue in the US, as is weather related delays.
There are absolutely circumstances where flights on different carriers end up on different itineraries. Sure, TS is one possible solution in the event of delays but so are processes that make dealing with delays out of the ongoing carrier's control more passenger friendly.
>that empty seat doesn't mean you can take less fuel
I was under the impression that the fuel load for each flight is calculated based on a formula that averages the weight for each passenger and their expected luggage. Fewer passengers than expected = more fuel left over after the flight.
If the 100-passenger airplane always had only 99 passengers, there would probably be noticeably lower fuel cost. But a 99-person airplane would be even cheaper.
Yep you're right, everyone employed by Southwest to research and implement this apparently crucial detail of airline travel has no idea what they are doing and totally fucked up.
I honestly think hacker news should implement a feature where your comment is read back to you slowly before you post. Or you have to retype it out twice. Something, anything, to give the blowhards on here pause before they bless us with their cited[0] wisdom.
I think if there's any airline that I'd give the benefit of the doubt for making a change like this, it would be SW. I can't imagine that a conversation around this issue ONLY came up for them after the (most recent) United fiasco.
Not all airlines overbook. Case in point, JetBlue, which as a practice does not overbook. United heavily overbooks, which is closely related to their practice of trying to offer the cheapest tickets on major competitive routes.
United lost over $2 billion last year. JetBlue made about $800 million in net income last year. So which strategy is "economically important" again?
Yes, that's what "standby" tickets basically are. Some airlines would (not sure it's still a thing) sell you a "standby" ticket for cheaper than a normal ticket, which meant you show up and wait at the gate until near departure time, and then if there are any seats available they let you on the plane. If not you have to wait for the next flight.
Yeah, but the problem is that most tickets are now, in effect, standby, but not labeled as such. Hence why the parent suggested making the status (and the option to upgrade) clearer and more readily available.
On Delta the Basic Economy (E) basically are standby since you can't select your seat and are only assigned one at check in (or at gate? can't remember). It's not even that much cheaper than regular main cabin.
According to the article, 4% of planes are overbooked. And if they are, <5 will usually be re-booked. Which is a low single-digit percentage of passengers. So your chance of being denied boarding is probably <0.1%. Wouldn't say that all tickets are standbye.
It's rather that thanks to the internet we know of stories like these. 10 years ago no one would've known.
I say that you're "effectively on standby" when there's a potential for the overbooking to keep you off the plane, which is the kind of risk that should be made explicit.
If you buy a higher fare class you won't get bumped -- or extremely rarely. I fly 60,000 miles per year and I have never been bumped. But I also don't buy bulk-fare consolidator tickets from travelocity-type sellers either.
Involuntary bumps are extremely rare statistically anyway.
While I generally agree -- Southwest is more of a budget airline, so it may not be subject to the same magnitude of loss from people missing/skipping their flights. I could see the United demographic having far more no-shows.
For those who aren't aware, airlines overbook because there are, on average, a significant number of people who miss or skip their flights. It's wasteful (and bad business) to leave those seats empty when instead, you can overbook and have an extremely low percentage of flights end up without enough seats.
Edit: as others mentioned, overbooking also allows airlines to be more flexible with refunds.
I understand the benefits of overbooking and I think it's an overreaction to stop it entirely.
But, couldn't they make the tradeoff more explicit and transparent for the customer? "This ticket allows you to be bumped, for which you will be paid $X. You can pay $Y more to make it a permanent you-own-this-slot-we're-not-f'ing-around reservation."
None of this, "oh, you didn't read and memorize the minutiae of overbooking policies and have to figure everything out minutes before the flight you thought you were going to be on!"
It already IS transparent. Every single ticket counter generally has some form of overbooking notice. -- at lesst they did when I flew through DFW a few weeks ago. I don't think anyone reading this is unaware of overbooking. It isn't some mysterious secret clause on a TOS page somewhere.
Overbooking is key for maximizing producer surplus. Whether or not that is economically advantageous to the economy as a whole is a different and more complex question.
But the notion that a company maximizing it's surplus is _always_ beneficial to consumers is tricky but more likely false within the confines of my limited understanding of contemporary economic theory.
It's necessary to maximize total surplus as well. If someone wanted to get on the plane, and they took off with a seat empty, that's a deadweight loss. The guy sitting next to an empty seat is a little happier, but it's pretty minimal.
United increasing the maximum bump fee completely solves the problem. Short of the apocalypse, you're going to find plenty of takers on any flight for $10k, even if it's a multiple day delay. The people who take the payoff are better off -- if they weren't, they wouldn't have taken the voluntary deal to fly later. Then the airline just needs to gather some statistics about how high the auctions go, and optimize the overbook rate to maximize their revenue. Everyone then is better off. In fact, on a lot of flights, you can probably find multiple people to bump with just a token payment -- they'd rather stay another day.
The other great thing about dynamic overbooking like this is that you never have to have a sold out flight, some of them are just really expensive. Someone who really needs to be on a plane walks up and plunks down $5000 or something, and the airline finds the least disrupted passenger, and pays him a few hundred. The bumped guy is happy to have the money, the last minute guy is happy to be on the plane, and the airline is happy to make $4700 extra.
This works even better with a name-your-price bump question on checkin for flights that are oversold. Even if it's nonbinding, the airline now knows that it has willing passengers to bump and can keep selling tickets.
This is why I left it vague...it really is a complicated issue with several angles. That...and my Bachelors only goes so far.
That said, I think you are conflating a couple of economic principles.
> It's necessary to maximize total surplus as well. If someone wanted to get on the plane, and they took off with a seat empty, that's a deadweight loss.
I think you are confusing utility [1] with surplus somewhat here. If someone wanted to get on and they took off with an empty seat that is a loss in Utility. If the airline turns around and sells that seat to the new person that increases Producer Surplus (and conversely decreasing Consumer Surplus) to increase Utility.
I don't think this situation (empty seat) is DWL in the traditional sense [2]. The traditional DWL scenario is when producers could supply more but don't because of pricing constraints or other artificial ceilings/floors. It is definitely a marketing allocation inefficiency which results in Utility loss and I am kind of interested to see if there are better quantifications of this loss...but it isn't a DWL in surplus technically.
> Everyone then is better off. In fact, on a lot of flights, you can probably find multiple people to bump with just a token payment -- they'd rather stay another day.
It's totally possible that everyone can be better off. The issue isn't about when everyone is better off...it's when the system breaks down and you have situations where no one wants to make that trade.
> The bumped guy is happy to have the money, the last minute guy is happy to be on the plane, and the airline is happy to make $4700 extra.
This is an example of maximizing producer surplus. In my mind it's a variation of Price Discimination [3]. And while this overall beneficial situation is plausible, it isn't really overbooking.
For me, overbooking is basically selling more goods than you actually have and playing on the statistics trying to bank on people not making the flight. If you sell 30 seats and 5 people don't show, you take off with 25 people on board. If you sell 35 seats and 5 don't show, you take off with a full flight and pocket some additional revenue from the 5 no shows.
Now there is a lot of gray area since most no shows are from missed connections and so some of that additional revenue can go back into the system to allow for cheaper cancellation fees as well as accommodating rescheduling costs for these missed connections. But in general I suspect that producer surplus increases more than the gain in consumer utility.
> This works even better with a name-your-price bump question on checkin for flights that are oversold. Even if it's nonbinding, the airline now knows that it has willing passengers to bump and can keep selling tickets.
This is again, Price Discrimination but now adding information imbalance in order to help the airline more effectively Price Discriminate thereby maximizing their Producer Surplus.
Technically the United situation isn't the result of overbooking anyway...but I still think the practice warrants scrutiny. United's recent announcement to increase payouts and mileage compensation is testament to the fact that there was Producer Surplus to spare.
Like I said, all of this is speculative discussion and there is a lot of gray area. It is also plausible that overbooking results in maximizing Utility and an even handed redistribution of benefits to both producers and consumers. But I believe Producers are driven by incentives...and they don't have strong incentives to share the benefits of overbooking altruistically back to the consumer.
I fly from SFO - PDX every other weekend and consistently, no matter how early I book, Virgin America flights are cheaper than United and Virgin doesn't overbook whereas United does. Generally speaking, if United were even slightly returning the added revenue from overbooking to the consumer shouldn't their flights be cheaper than non-overbooked competitors?
Again this observation and subsequent question is anecdotal and broad...but I like it for FFT.
There is a difference between Surplus and Utility. I think there is a case to be made that this practice increases Consumer Utility and trades it somewhat for Producer Surplus...but generally speaking, Price Discrimination increases Producer Surplus (thereby reducing Consumer Surplus).
Price discrimination does not necessarily reduce consumer surplus. In this case in particular, there is a large surplus being created by the fact that planes fly more full, and I don't know any reason to think that consumers don't capture much or most of this .
I would be interested to see how you would graph price discrimination not reducing consumer surplus. Price Discrimination in itself is about redistributing surplus from consumer to producer.
> a large surplus being created by the fact that planes fly more full
I think you are conflating surplus with utility. Full flights have more utility than not full flights...but the market surplus available is unaltered.
Maximizing utility is not a bad goal to have...but it is distinct from Consumer and Producer surplus.
> I don't know any reason to think that consumers don't capture much or most of this .
Because in general most of the surplus is captured by the Producer and the producer doesn't have any incentive to redistribute it back to the consumer. Not having an incentive doesn't mean they don't, I am just pointing out airlines aren't incentivized to reduce the seat cost of a late purchaser by the amount of the cancellation fee for the person that cancelled their flight...or even offer the seat at the price the original person paid.
Anecdotally (as mentioned in my other comment) we can compare United to Vrigin. Let's assume that a whole bunch of "surplus" is created by overbooking. United overbooks, Virgin does not. If the consumer captures most of this than United flights should be cheaper than Virgin ones in almost every way (ticket price, bag check fees, cancellation provisions, etc.). I fly from SFO-PDX twice a month. Virgin America consistently has lower priced fares than United.
This is observational and anecdotal...but the result casts doubt on the hypothesis that the consumer captures most of the surplus generated by overbooking. Admittedly, I wouldn't expect the consumer to do so from a purely theoretical stand point .
> I think you are conflating surplus with utility. Full flights have more utility than not full flights...but the market surplus available is unaltered.
Utility is quantity that is measured with respect to an individual (subjective), whereas surplus is an objective measure. This is but one difference, and I am not conflating them. Fuller flights (less dead-weight loss) means there is more surplus to go around, and which will in general be divided up between the buyers and sellers.
I could say similar things about your understanding of price discrimination. My guess is that most of your confusion comes from thinking in terms of an idealized model where prices are dominated by marginal costs, whereas fixed costs for planes are very large.
This can all be inferred from the definitions you link to, you just haven't internalized them.
The $10,000 change at United depends a lot on the implementation.
For instance, if they continue to have rapidly-expiring vouchers, most travelers will not have any practical way to use the whole reward. Also, the airline could add conditions to vouchers such as a voucher that is completely “used up” by a ticket purchase even for tickets that cost less than the voucher amount. There are lots of ways they can try to create good press by technically giving you something that is hard to use.
The first claim in your first link... exactly what I expected to find: "By overbooking it actually does help keep the fares down"
Money. Low low prices. That's why people that have paid for tickets are getting beat up and kicked off planes.
In all sorts of other aspects of our world we're expected to gladly accept higher costs as the price of our "values." Higher energy costs, mandates for various types of insurance, taxes to fund one agenda or another and on and on. Perhaps you haven't considered this yet but delivering the seat that a paying customer has paid for and not subjecting any of these otherwise compliant paying customers to the possibility of being thrown off a plane --- by martial force if necessary --- is a "value" we might aspire to.
And if that means maybe you are a little more careful about just how often you bop across the continent in a 600 mph jet because the cost is a little higher, or you can't play ticket games with as much abandon as you're used to well... perhaps that isn't really the end of the world.
The only time I experience overbooking is when travelling in the US. I have never had someone come on a flight, or announce at the gate, that people must give up their seat and offer compo (something I have experienced on probably 20-30% of my domestic US flights).
If it gets really high, I imagine passengers would do their best to drive it up. Would you volunteer if the offer was $2k out of a $100k vs a $10k pool? It defies classic game theory to an extent, but humans like to roll the dice.
You're going to collude with ~100 random strangers to prevent anyone else from taking the offer until it gets to the level you want it to get to, and then agree to split it somehow? Ha ha ha.
Last year when I was flying back from LIH this happened but on a very small scale. These two guys both were joking about just driving the price up and so they did, all the way to the max Guy got a free hotel stay and food as well since there was only one flight per day to LAX.
There's nothing to suggest the model is overly simplistic. Good luck convincing 100+ people to collude on those shenanigans at their own personal loss.
Someone will eventually defect, probably far lower than $10k. If it becomes a very high cost for airlines, they will simply price that externality into their calculation for double booking.
I have a question for the HN crowd: what would happen if airlines sold their seats at a fixed price and let customers resell them?
For the sake of the discussion let's assume that all seats are in the same class (ie: economy) - there are no business / first class seats.
So the idea would be:
1. All seats are sold at the same price, no matter how early or how late you buy. The price covers all the costs associated with running this flight, given an expected "occupancy rate" (sorry not the right term).
2. All tickets are non-refundable. If a customer changes their mind, they use the airline app to put their ticket up for sale. They can try to sell the ticket at the same price, or at a discount (or maybe even at a higher price?)
3. Airline charges a small fee (something like $10) every time the ticket is transferred to another person.
Since this has never been done - to my knowledge - there must be something wrong with it. But I wonder what it is?
The problem here is really selling all seats for the same price.
Airlines know that some customers are price-sensitive, e.g. those planning a trip months in advance, and some customers less so, e.g. business people flying out to a last-minute meeting.
So if they just sold all the tickets for the "average" price, they simply wouldn't get as many customers, because all the price-sensitive customers would go to their competitors still doing things the old way, and there's no guarantee that they will attract enough of the less price-sensitive customers, unless they lowered prices even more than the previous average, and if they actually make less than the current average price, this is simply not a viable plan.
In addition, since the customers you have left tend to book later, you won't know for sure if you're filling up your plane or not until closer to the departure date, and that could potentially wreak havoc when allocating planes to routes.
> Since this has never been done - to my knowledge - there must be something wrong with it. But I wonder what it is?
Sure it has. Look at concert tickets. You are basically talking about introducing the scalper into the airline ticket market.
Try to get any tickets to a major act at anything approaching face price - especially for the prime seats. You can't do it. The scalpers always beat you to it.
Now imagine, the only seats left after the scalpers get through are the tickets in the way back.
Now lets just throw in the mix handling flight changes. Because from the airlines' perspective every flight is full. You could never book through the airline....
The challenge is whether you are fundamentally buying a seat on a particular plane, or transit from point A to point B. Almost all simple solutions re: air travel break down once connections get involved.
Also breaks down for equipment swaps. If we view seats on a given flight as assets that can be resold, there's a lot of ways that normal (even understandable) airline operations can "devalue" those assets, and sorting that out legally would be an amazing mess.
Pretty much any American airline that provides "better" service generally does something along the lines of having a very limited coverage map and a more-or-less a uniform fleet.
Most problems with air travel are "last mile" problems, with "last mile" is defined as flying to a location that does not have a critical mass of service to really be a "full service" city.
The issue I have with that is then these middle-men buyers would step in like they do for NFL and concert tickets, buying up all/most supply almost immediately when it becomes available for any reasonably popular flight. Then they charge double the price, and there's not much one could do other than pay double or not fly.
>what would happen if airlines sold their seats at a fixed price and let customers resell them?
Buying airline tickets involves multiple checks to minimize the possibility of the buyer blowing up the plane, and resale on the secondary market will weaken those
You used to be able to do this in the long-ago; you'd even see people selling tickets in the classified section of the newspaper (kids, ask your parents).
The problem is, the airlines would rather make more money by selling the new person the ticket at an inflated price (presumably if they cared enough to save money, they'd have bought in advance), and from you as well in ticket change fees. This was sold to us as a post-9/11 fix to ensure terrorists didn't buy tickets, but I doubt that was the underlying reason.
So, stubhub for plane tickets?
I think there's some difficulty around changing the name that's on a ticket, but other than that, it seems like a neat idea.
I think there are a lot of rules around reselling tickets post-911. I'm sure you could find a way to execute it while accounting for the new regulations, but for most airlines i don't think its worth it. They also make money off of a fluctuating ticket price or they wouldn't have them.
The way I understand it, if airlines were to offer fixed-price tickets with no class distinction, then there wouldn't be almost any flights - i.e. there's no point at which, for a typical flight, number of tickets bought * fixed price >= costs of fueling a jet and paying the crew.
Exactly, even when you add in the transfer fee proposed here. Any rational ticket holder wouldn't sell their ticket for less than [what they paid for it] + [the transfer fee], and that means the airline let someone else get money they could have got if they had just sold the ticket directly to the second person.
The only scenario where this would be useful is if the the original purchaser is willing to sell their ticket at a loss for some reason, e.g. if they know they can't make it and just need to minimize their losses. But at least Southwest airlines already has a customer-friendly solution here, they will credit you the full price of the ticket for a future flight, for most tickets at least.
Original purchasers sell their tickets at a loss back to the airline all the time (for credit). It's called canceling your flight because your plans change for some reason.
As you point out, Southwest lets you do this for free. And JetBlue is something pretty reasonable as I recall. Unfortunately most of the fee-addicted airlines have really ratcheted up their cancellation fee however.
I think my favorite solution to this is to sell "guaranteed" and "standby" tickets. Sell full plane capacity of "guaranteed", and then any ticket after that is bumpable/standby. You could potentially switch to sell "bumpable" tickets earlier, too, at a discount.
If the airline is flying ~8 flights a day from SFO-SEA, I'd be fine in many cases paying x% less for y% risk of being bumped one or two flights later.
People are price-sensitive (will always buy the cheapest option) and then complain when they don't get the expensive service. It is incredibly difficult to explain to someone: "You are not getting to your daughter's wedding because reasons." ...and nobody wins that argument when it happens.
Apart from removing over-booking completely, the Delta "pre-blind-auction" is the best choice. All tickets are sold at their regular prices (no random discounts), and day-of people are able to decide if they are willing to be bumped (ideally: by how long) as well as the price it would take.
There's a huge difference in "I'll take $1000 to be bumped to a later flight today" and "I'll take $1000 to be bumped to a flight that's two days later" (depending on the route, that is something that can happen).
It's a delicate balance between customer service, incentives, and profit, and "guaranteed v. standby" pretty clearly fails.
The pricing nature of an airline seat is such that the closer to the flight time the more expensive the ticket is. Consumers have been forced to pay more under this model for the longest time. Buying back seats due to overbooking should follow the same rule. Buying back seats at the time of flight departure should cost the most. If airlines follow their own pricing model, they won't have a problem. It's only when they try to bend the rules and use forces to get back those seats that people got upset.
Between San Francisco and Los Angeles, Southwest has a lot of flights on their schedule, but if one isn't full they'll just cancel it. Often all the midday flights will be cancelled. Technically, this isn't overbooking, but it feels like it.
Seems odd that the airlines keep talking about overbooking but while yes overbooking is an issue, it's not this issue. This issue is that UA wasn't overbooked but rather allowed someone to get kicked off who was already boarded because late staff wanted to catch a free ride.
JetBlue doesn't overbook, but they have removed passengers already boarded who have not violated any rules, other than they wanted to accommodate other passenger(s).
Overbooking is an issue, and they should either stop overbooking (though that means potentially higher ticket costs for everyone), or when they do overbook make sure they don't start boarding everyone until someone has agreed to not fly and offer the appropriate amount to get someone not to fly, such as the max amount or cash rather than a voucher. But once again, that's a different discussion, and SW and others should focus on the bigger issue at hand that UA violated.
They started to use them, that's not the difference. The main difference is that Ryanair doesn't allow connecting flights. So it's always your fault if you miss the flight, never Ryanair's.
This is stupid. As a passenger I benefit from overbooking because it means lower cost per ticket and the option to get free flights out of an overbooking scenario when I have a flexible schedule.
For anyone looking to switch airlines amidst all the news lately, I can't recommend Southwest more highly. Flew them out of Chicago for business the last two years -- their rewards, no-fee rebooking, etc. literally changed our lives. Wrote a bit about it here:
Southwest already allows cancelling your ticket up to the last minute for a future credit so presumably most people that will miss the flight do so already. I wouldn't expect this impacts their fares much, if at all.
Now if United, AA, or other large airlines that don't offer cancellation did this, that'd be another story.
>Southwest already allows cancelling your ticket up to the last minute for a future credit so presumably most people that will miss the flight do so already. I wouldn't expect this impacts their fares much, if at all.
One reason you'd let people cancel at the last minute is because you overbook flights. If you have 100 seats, book 105 people and then 4 cancel, that cancellation didn't cost anything.
If you book 100 people and turned away other passengers, then 4 people cancel at the last minute, now you lost 4 sales.
Southwest already often isn't the cheapest, or even close to the cheapest, fare on many routes. And whether this causes an increase in fares will mostly depend on the existing load factor. If they already had planes mostly going out with a few empty seats, they don't necessarily have to change anything.
They are currently in the process of moving to a more flexible scheduling system (until now they could only have a weekday and weekend schedule) which should allow them to have more profitable flights (as they dont need to flt an empty plane for 3 days if only 2 days are high volume).
I've flown very often with Southwest, and have rarely seen a completely full plane. The fullest I've seen still has one or two middle seats free at the very back of the plane.
I can't even remember the last time I was on a flight (Southwest included) that wasn't at capacity. I'm sure it depends on where you're flying to and from though.
The time of day matters though. Early morning departures, in my experience - especially on weekends, are almost always partially empty. Last week I flew on a flight that left at 5:55 AM that was less than half full (25 people on a regional jet). Last December I flew on a southwest flight that left at 7:45 AM that was about half full (probably 60 people on a 737).
I don't see how it could. Only a handful of people on any one flight are ever bumped, on average. The bumped are normally put on other flights. The only 'win' is where someone doesn't show and also doesn't fly or get a refund. I cannot see how it is more than 1 or 2% of revenue.
That said, perhaps I am dead wrong and overbooking is being used for darker purposes. Perhaps they are deliberately overselling flights at peak times knowing that they will push bumped customers to later less-full flights. If that is being done knowingly, that's a bait-and-switch operation that might be worth more than a couple percent.
Flights are usually overbooked by around 5-10 passengers a flight, and is well-calculated based on the expected number of no-shows.* Depending on the ticket price, this can mean somewhere between $1-5k a flight, which can be a pretty substantial difference for a lower margin flight.
No, there's no benefit to deliberately overbooking to fill a later flight given that bumped passengers are compensated. At best they're giving away a seat on that later flight for ~free, and at worst they're eating the cost of another paying passenger that can't book that seat anymore.
*source: a friend working in revenue mgmt at United
Given the reason that airlines employ overselling, you would have to expect that it would impact the airline profitability, which could in theory lead to an increase in ticket prices. The bet SWA is making is that the revenue gain from overbooking is smaller than the consequential losses from bad publicity. Just think about the downside to United based on their share price alone, and now the increased focus that is on the 'good conduct' w.r.t bumping passengers in the light of it.
The maths of overbooking is that if you have 100 bookings on a 100 seater flight and 97 passengers show up then perhaps you could have sold 103 tickets and still not have to bump anyone. In this case you have increased revenue by 3% without an increase in costs. This can either manifest itself in an increased profit or, on a competitive route, lower ticket prices.
Of course, there will be variations in the number of people who show up and this can lead to more passengers arriving at the airport than there are seats on the plane. In this case, there is a cost to the airline, and perhaps in this big optimisation problem there are some airlines that see a bigger payoff in overbooking and bumping than avoiding overselling and minimising bumping.
I don't know about budget airlines, but full service you might expect somewhere in the region of 5% overselling. For budget, I would expect it to be higher as tickets are cheaper so passengers have less to lose by not showing up.
Nonetheless, even if it were a 1-2% increase in revenue, it would have a much higher impact on profitability. On a budget route, you could have $500 profit per flight, say a couple of dollars per passenger. Every extra ticket that gets sold that the airline doesn't have to pay costs for is a big deal, and part of the optimisations that allow airlines to offer lower fares.
I don't want to fly on an airline with profit margins so slim. If a few hundred dollars one way or another is all that matters, I don't want to think about the decisions going on behind the scenes. These are extraordinarily expensive machines filled with hundreds of very human humans. You need more room for contingencies. I'd rather pay a few dollars more if that meant flying on a plane that wasn't constantly skirting the edge of profitability, one where employees have room to take reasonable and safe decisions.
>I don't want to fly on an airline with profit margins so slim.
You should probably consider switching to Greyhound in that case. The last few years have generally been good ones for commercial aviation but, historically, the industry lost money as whole since its inception.
One of the problems is that you may be willing to pay a few dollars more but most will just go with whatever is cheapest.
It's not about how many were bumped. It's about over selling seats so that when people on refundable tickets don't show the plane doesn't leave with more empty seats.
Margins in air travel are small, an extra seat sold on every flight is a huge difference, this will definitely affect revenues and profits, and will likely lead to slightly higher ticket prices. Until the race to the bottom returns and they start overbooking again.
Southwest doesn't do this, but some other low-cost carriers do (especially in Asia): they purposefully cancel flights and offer only credit vouchers as compensation instead of re-routing passengers. Many travelers end up buying same-day tickets at considerably higher expense, or not ending up in a position to use those vouchers (which is the same as the airline keeping the money).
Well, if it's a 1-2% decrease in revenue there might be a 1-2% price increase. I thought the margins for US airlines were incredibly small (could be wrong, I'm no expert).
They give flight vouchers, but they often have other direct expenses like hotel accommodations and food.
The last time I was bumped it was ~$600 airline voucher, transportation to/from a ski resort (a perk of being bumped in the mountains!), a room for the night and a good meal voucher to cover dinner and breakfast the next day. Plus they still flew me out the next day and took care of rebooking the connection on another airline. I don't know exactly how much it cost them, but I made out like a bandit.
I can better that. In the early 90s I (12) and a schoolmate (17) were flying vancouver-toronto-london with BA. I as an unaccompanied minor, him no so. The toronto-london flight was delayed overnight due to mechanical issues. They probably wouldn't have helped my friend so much but when I mentioned there was an older student on the flight from the same school they tracked him down. We got two hotel rooms and, this sounds strange but true, a stewardess. She took us to visit the University of Toronto, to which my friend was applying, and then the CN tower before our flight the following night.
In the US at least, they are now considered as a commodity. But the consolidation of major carriers (American with US Air, United with Continental) may have changed that equation.
There is no "win" for the consumer here, same as when airlines decide to re-provide free food or free checked bags - these things get baked into the price of tickets, nothing is free. They are marketing. Rather than spend a million on ads, they spend a million on less efficient flights (by not overbooking), but hopefully getting more customers than ads would have gotten given this is a hot-button topic.
Usually when you miss your flight you get credit. So non-refundable just means you can't get cash. Most times you still can use the price of the ticket on another flight in my experience.
I've seen a number of overbooked southwest flights and people can't line up fast enough for a $300 flight credit + next available flight. I can't imagine it ever going over $1k.
IMHO if any airline should continue overbooking it's SWA. No first class, no assigned seats. Most of the flights are short range with multiple flights options every day. You're much more likely to agree to switch to the next flight if it means only waiting 2hrs.