Saturday, November 6, 2010

Directive 261: why the airlines should pay for hotel (and why they shouldn't)

Here I discuss the issues related to occasional cancellations, not caused by the airlines. These arguments may not be valid if the airline itself is the source of the problem, or in case of massive air traffic disruptions, such as the ash cloud of 2010.

By Jnpet [GFDL or CC-BY-SA-3.0], from Wikimedia Commons

There are two reasons to require airlines to provide food and lodging -- risk aversion and economies of scale. Humans do not like risk. They hate to change their plans in an unexpected and expensive way like when the flight is cancelled because of bad weather or industrial dispute. Obviously, the airlines hate to pay these costs as well. They simply collect the money from the passengers themselves in form of more expensive airfares. Essentially, the airlines act as insurance providers, charging passengers a little more for their flights, and providing food and shelter when needed. As airlines operate many flights, delays and cancellations occur quite often, and it is relatively easy for them to collect and maintain related funds.

Large firms can handle occasional disruptions lot easier than individual passengers. Hence it makes sense to shift more of the risk to the firms. Second, as airlines have a base or partner in the airport, it is easier for them to provide the hotel. Imagine yourself stranded in an airport far away with little money, extremely expensive cellphone prices and no idea about the hotels. How would you get to an affordable one? Airlines could easily (eventually through airports) make an agreement with a number of hotels and taxi companies. Even more, they have access to phones, internet and office facilities, which tremendously simplifies booking. It costs a lot less for them to do the actual bookings in case of cancellations.


The arguments above work best if all passengers are similar. But they differ. First, people have inherently different tendency to take risks. Second, your risk aversion depends on the circumstances. Your outbound flight was delayed? Bad, but not a big deal. Just go home and sleep till morning. You are probably not that interested in a hotel just a few blocks away from your home. However, if this happens to be your flight home, the situation is different. Third, people value lodging differently. A backpacker may find it completely acceptable to spend a few nights in the airport (in exchange for cheaper airfare), while others are willing to pay a lot for a good hotel.


The current directive 261 requires the airlines to offer a one-size-fits-all compulsory insurance. But sometimes you may rather want to get cheaper tickets and take the risk, another time you may want to pay even more to get better provision. The current directive does not allow for this kind of flexibility. Fortunately, there is an easy solution -- make the insurance voluntary, an explicit choice with corresponding price tag, while buying the tickets.

Does this mean that current directive 261 should be replaced by a voluntary insurance? Maybe. It depends on how different the passengers typically are, and whether they actually understand the risks and make the right choices. But it might be a good idea to experiment with this option.

This story is inspired by discussions with Idir Laurent Khiar. Here I use the opportunity to acknowledge his role.


  1. We would offer one more argument -- moral hazard. Although you talk about cancellations that are not caused by airlines, airlines still have some control over the circumstances, and therefore the probability of disruptions: they choose frequency of flights, the quality of service providers, etc. Their efforts towards avoiding cancellations are not generally observable and therefore requiring them to cover the total social costs may lead to more efficient allocation of resources. Similarly, if we only take into account risk aversion and economies of scale, it would not be reasonable to punish the customers who arrive to airport too late, but otherwise the negative effects due to moral hazard would probably outweigh possible benefits.

  2. Agree. Just I did not want to incorporate the airlines role as the story is long anyway...

    Additionally, there may be long and costly legal processes, trying to establish whether airlines were the culprit of the particular technical problem (versus producer/maintenance/reparation/ATC ...).

    Passengers arriving late: yes. Someone suggested to make last-minute check-in costly, to discourage coming late, but still allow for efficient allocation.

  3. I remember our discussion – glad to see that it led to an article! Here, a longer comment: A clarification and a broader view in terms of economic efficiency.

    First, Directive 261 provides not for insurance. It simply clarifies a compensation matter. This is important. It clearly limits the scope of the question guiding your article to insurance relevant aspects, such as weather events, volcanic ash, or terrorist attacks - not covered by the Directive. These are insurance issues as a probability of occurrence can be assigned to them. Should airlines compensate their customers for additional expenses in case of such events, such as hotel bookings etc.? What does it mean in terms of economic efficiency?

    As we deal with matters which can be quantified in terms of probability, the issue at stake goes beyond simply covering additional costs. It includes for the involved actors prior to a flight to assess the risk of occurrence and take the necessary steps to cope with it (e.g. enough cash to spend on hotel to stay overnight). The information advantage and economies of scale clearly point towards airlines to be tasked providing this insurance (taking outsourcing to insurers into account). The same is true for establishing necessary capital reserves to be used in case of an event. Both is done economic more efficiently by the airlines (or usually insurance companies) than by individual customers. However, the related costs occurred prior to the flight disruption may easily outweigh the costs occurred immediately after the disruption, such as for hotels, taxi etc. It explains why a company may generously pay for an expensive taxi trip or hotel expenses without legal obligation. Economic efficiency explains why airlines should provide for insurance but also why they do not on a general base.

    As mentioned, Directive 261 does not force airlines to compensate for events outlined above (see Art. 5). Such events, commonly referred to as force majeure, fall out of scope. Those are events which could not have been avoided by an airline. Nevertheless, it would be possible to provide customers insurance – but only on a selective basis, e.g. volcanic ash, terrorism etc.).

    Again, economic efficiency provides the answer why Directive 261 does not intend to make it mandatory for airlines to provide such insurances. Given the low probability of such events, establishing huge capital reserves may be more costly over time than expenses that result if case of an event. Finally, with all insurances it will be the customer who will have to foot the bill. Considering the overall economic efficiency of such an insurance (not to mention complexity) lawmakers decided that airlines should not pay for costs due to disruptions as outlined above.