United Airlines‘ mistreatment of a passenger last week in Chicago had nothing to do with overbooking, but commentary on the issue of overbooking is suddenly all the rage. The fiasco in Chicago began when four United employees arrived at the gate after a flight to Louisville had boarded. The flight was not overbooked, just full, but the employees needed to get to Louisville. United decided to “bump” four passengers to clear seats for the employees. They used an algorithm to select four passengers to be bumped based on factors like lowest-fare-paid and latest purchase. The four passengers were offered vouchers for a later flight and a free hotel night in Chicago. Three of the four agreed, but the fourth refused to budge. United enlisted the help of Chicago airport security officers, who dragged the unwilling victim off the flight, bloodying him in the process. It was a terrible day for United‘s public relations, and the airline will probably end up paying an expensive out-of-court settlement to the mistreated passenger.

Putting the unfortunate Chicago affair aside, is over-booking a big problem? Airlines always have cancellations, so they overbook in order to keep the seats filled. That means higher revenue and reduced costs on a per passenger basis. Passengers are rarely bumped from flights involuntarily: about 0.005% in the fourth quarter of 2016, according to the U.S. Department of Transportation. “Voluntarily denied boardings” are much higher: about 0.06%. Both of these figures seem remarkably low as “error rates”, in a manner of speaking.

Issues like the one in Chicago do not arise under normal circumstances because “bumps” are usually resolved before boarding takes place, albeit not always to everyone’s satisfaction. Still, if airlines were permitted (and willing) to bid sufficiently high rates of compensation to bumped ticket-holders, there would be no controversy at all. All denied boardings would be voluntary. There are a few other complexities surrounding the rules for compensation, which depend on estimates of the extra time necessary for a bumped traveler to reach their final destination. If less than an extra hour, for example, then no compensation is required. In other circumstances, the maximum compensation level allowed by the government is $1,300. These limits can create an impasse if a passenger is unwilling to accept the offer (or non-offer when only an hour is at stake). The only way out for the airline, in that case, is an outright taking of the passenger’s boarding rights. Of course, this possibility is undoubtedly in the airline’s “fine print” at the time of the original purchase.

No cap on a bumped traveler’s compensation was anticipated when economist Julian Simon first proposed such a scheme in 1968:

“The solution is simple. All that need happen when there is overbooking is that an airline agent distributes among the ticket-holders an envelope and a bid form, instructing each person to write down the lowest sum of money he is willing to accept in return for waiting for the next flight. The lowest bidder is paid in cash and given a ticket for the next flight. All other passengers board the plane and complete the flight to their destination.“

Today’s system is a simplified version of Simon’s suggestion, and somewhat bastardized, given the federal caps on compensation. If the caps were eliminated without other offsetting rule changes, would the airlines raise their bids sufficiently to eliminate most involuntary bumps? There would certainly be pressure to do so. Of course, the airlines already get to keep the fares paid on no-shows if they are non-refundable tickets.

John Cochrane makes another suggestion: limit ticket sales to the number of seats on the plane and allow a secondary market in tickets to exist, just as resale markets exist for concert and sports tickets. Bumps would be a thing of the past, or at least they would all be voluntary and arranged for mutual gain by the buyers and sellers. Some say that peculiarities of the airline industry argue that the airlines themselves would have to manage any resale market in their own tickets (see the comments on Cochrane’s post). That includes security issues, tickets with special accommodations for disabilities, meals, or children, handling transfers of frequent flier miles along with the tickets, and senior discounts.

Conceivably, trades on such a market could take place right up to the moment before the doors are closed on the plane. Buyers would still have to go through security, however, and you need a valid boarding pass to get through security. That might limit the ability of the market to clear in the final moments before departure: potential buyers would simply not be on hand.  Only those already through security, on layovers, or attempting to rebook on the concourse  could participate without changes in the security rules. Perhaps this gap could be minimized if last-minute buyers qualified for TSA pre-check. Also, with the airline’s cooperation, electronic boarding passes must be made changeable so that the new passenger’s name would match his or her identification. Clearly, the airlines would have to be active participants in arranging these trades, but a third-party platform for conducting trades is not out-of the question.

Could other concerns about secondary trading be resolved ion a third-party platform? Probably, but again, solutions would require participation by the airlines. Trading miles along with the ticket could be made optional (after all, the miles would have a market value), but the trade of miles would have to be recorded by the airline. The tickets themselves could trade just as they were sold originally by the airline, whether the accommodations are still necessary or not. The transfer of a discounted ticket might obligate the buyer to pay the airline a sum equal to the discount unless they qualified under the same discount program. All of these problems could be resolved.

Would the airlines want a secondary market in their tickets? Probably not. If there are gains to be made on resale, they would rather capture as much of it as they possibly can. The federal caps on compensation to bumped fliers give the airlines a break in that regard, and they should be eliminated in the interests of consumer welfare. Let’s face it, the airlines know the that a seat on an over-booked flight is a scarce resource; the owner (the original ticker buyer) should be paid fair market value if the airline wants to take their ticket for someone else. Airlines must increase their bids until the market clears, which means that fliers would never be bumped involuntarily. A secondary market in tickets, however, would obviate the practice of over-booking and allow fliers to capture the gain in exchange for surrendering their ticket. Once purchased, it belongs to them.