Business Columns & Blogs

The economic lessons of the United Airlines passenger who was dragged off the plane

The kerfluffle over United Airlines having a passenger dragged from a flight is an interesting case study for managerial-econ teachers like me.

The case exemplifies the “principal-agent problem” at multiple levels. This is one failure of conditions necessary for a market to reach optimal efficiency. It occurs when an employee, the “agent,” should work in the interest of some “principal” such an owner, but then, due to badly designed incentives, works toward another end.

Having an already-boarded passenger dragged from a packed jetliner did not benefit principals, the people owning UAL stock. Yet some employee, an agent, felt compelled to take those actions.

The incentives in the incident are not clear. Some gate supervisor had to find seats on a full plane at the last minute for a flight crew that needed to reach St Louis. Failure to accomplish this by this employee would mean not performing his job..

We don’t know all options open to UAL reps. First they offered compensation to volunteers to take another flight. No one bit. Critics, including late-night comedians, wondered why the airline didn’t raise from $800 to $1,200 or $1,400 or whatever. Why not throw in hotel and meal vouchers? Or offer cash rather than vouchers?

One need not have a doctorate in marketing to know that police dragging a passenger would create bad brand identity. UAL was not even up to the $1,350 cap on what the law requires carriers to pay. Perhaps the supervisor faced rules limiting his flexibility.

These facts involve imperfect information and transaction costs. The employee may not have had the authority to offer more or no way to get authorization from higher managers. Employees who deal directly with passengers need a clear protocol to follow, and that cannot account for all eventualities. It all has an internal logic.

The biggest imperfection in information was the UAL rep’s failure to anticipate how calling police would play out. The rep probably had “We’ll see who is in charge” impulses, but he also probably thought, based on experience, that a defiant passenger confronted by police would leave quietly. The passenger didn’t. Passengers were outraged when he was dragged off the plane. So was the public.

In idealized perfect competition, travelers now would avoid United and fly other carriers. There would be as many sellers of flights as of doughnuts. Or, if one wanted to fly United, one could bargain with the ticket agent to remove the airline’s right to bounce you.

We don’t have perfect competition. Any single passenger has no bargaining power with any carrier. Air travel is competitive, but is far from perfect competition. It is oligopoly, a market dominated by a few large firms, each of which has overwhelming power over any customer. Even without overt collusion, oligopolistic firms adopt common policies on “terms of carriage,” so you gain no advantage in legal rights by spurning United for other carriers.

When the assumptions of perfect competition are not met, the outcome is not necessarily efficient. Action by government may improve efficiency, though that is not automatic. Decades ago, carriers had complete freedom to overbook and bump passengers as they saw fit. Customer outrage was impotent against the firms, but customers did gain political power. So we now have laws mandating compensation up to $1,350 and requirements for information. This helped things somewhat, both in efficiency and fairness.

Improving information through mandatory disclosures is a time-worn response to asymmetric information-market failures. But anyone who has closed on a mortgage knows that pages of fine print become meaningless. Information still is lacking.

Airlines are correct that overbooking lowers the average cost per seat. Once a flight is going, the marginal cost of carrying one more body in an otherwise empty seat is near zero. Average costs depend on the total cost divided by passengers. Having higher load factors lowers average cost. You can raise these factors by overbooking. Occasional bumping costs far less than the revenue gained by filling more seats.

Airlines argue that in a price-competitive sector, cost savings are passed to fliers. Some six passengers per 100,000 purportedly are bumped involuntarily. Millions benefit from cheaper tickets, carriers argue. Only a few bumped do not receive satisfactory compensation.

Skeptics argue that competition is not all that great. When fuel costs fell, carriers did not cut ticket prices, but rather engaged in share-buybacks that benefited stock holders.

Economics-savvy observers noted an element of behavioral economics in the incident: the endowment effect. In my column last week, I described blacksmithing tools that I would not have paid $100 for, but now that I have them, I would not sell for three times that. Once we have something, we value it more highly than before we got it.

The same is true for airplane seats. Once people are boarded and in their seats, they are angered much more by being told to get up and off than by being prevented from boarding in the first place.

St. Paul economist and writer Edward Lotterman can be reached at