Research
Class Actions: An Arbitrary Arrest
Arbitration clauses limit consumers’ ability to pursue collective actions.
Skiplagging occurs when a consumer purchases a multi-stop flight and disembarks at the stopping point instead of completing the entire journey, to save money.
For example, suppose a passenger would like to travel from New York City to Austin. A non-stop flight from New York City to Austin might cost $400. However, a flight from New York City to Denver with a layover in Austin might cost $200. The passenger may be motivated to book the cheaper flight and disembark at Austin, saving $200.
Skiplagging is prohibited by most airlines in the U.S. For example, American Airlines’ Conditions of Carriage prohibit “purchasing a ticket without intending to fly all flights to gain lower fares (hidden city ticketing).”1 Many airlines impose penalties for engaging in skiplagging, which include banning the customer from the airline, canceling the rest of the roundtrip ticket, refusing to let the passenger fly and check bags, and charging the passenger for what the ticket would have cost if they hadn’t skiplagged. Airlines argue that they lose revenue when passengers skiplag. If a passenger does not board the connecting flight, there will be an empty seat on the second leg of the trip, which the airline could otherwise have sold (albeit for the second time).2
In the U.S., airlines have objected to an innovative intermediary service called Skiplagged – a flight search engine that uses an algorithm to allow users to locate and book cheaper ‘skiplag’ tickets. Skiplagged explains that “Our goal is to empower consumers to use their buying power however they please. We exist to just find the best prices for you.”3 This intermediary therefore assists customers to purchase better value tickets. It plays the same role as rail ticket intermediaries such as Trainline that invest in algorithms that discover and sell cheaper split-fare tickets that are sold by rail companies in the UK and Europe. In response, some airlines – including American Airlines – have filed lawsuits against Skiplagged, alleging that it employs “unauthorized and deceptive” ticketing practices.4 In Europe, by contrast – many low-cost airlines (such as Ryanair for example) do not prohibit skiplagging.
Academics estimate that HCT opportunities arise on approximately 15% of US flights.5 This is consistent with a 2001 report from the U.S. Government Accountability Office, which found that HCT opportunities arise 17% of the time.6
The extent to which passengers exploit HCT opportunities when they arise is, however, unclear. Researchers found that some passengers are taking advantage of skiplagging opportunities, but noted that this usually involves purchasing two separate one-way tickets, given the risk that “failure to show up for the second leg of the outbound portion of the trip typically results in cancellation of the rest of the roundtrip ticket.”7 Other consumers may end up spending more by purchasing the higher-priced non-stop flight, so as not to violate the conditions of carriage.
Skiplagging also depends on route competition and primarily occurs on full-service carriers that operate large hub-and-spoke networks (American, Delta, and United).8 These airlines account for the majority of HCT.9 A study found that “an additional nonstop carrier on route A-C increases the likelihood of HCT by 1.6%-3.6% while an additional nonstop carrier on route A-B decreases the likelihood of HCT by 3.5%.”10 The study describes how, “by controlling a large fraction of flights and gates at their hubs, carriers are also able to exercise market power and charge a “hub premium” to passengers who originate or terminate their trips at the hub.”11
We examine two possible theories of harm in relation to skiplagging: (1) exclusion of a market intermediary; and (2) an unusual tying claim.
By banning skiplagging and seeking to exclude through legal action the intermediary Skiplagged, airlines are excluding a rival ticket retailer that imposes a constraint on their price setting. Skiplagged’s purpose is to increase transparency in the market and help consumers to locate the “best prices”. Therefore, using monopoly power to exclude it can be expected to lead to higher prices for consumers. Given the restrictions it faces, Skiplagged is not yet widely used. But if it were not suppressed by the airlines, the service it offers is an innovation that should incentivize rival intermediaries to invest in similar algorithms to improve their price competitiveness.
Notably, there have been parallels to this conduct outside the United States. In 2023, the German Competition Authority found that Deutsche Bahn – Germany’s largest rail company – abused its market power by withholding information, such as traffic data, from rival websites including Trainline.12 The same year, the European Commission commenced an investigation into Renfe – Spain’s state-owned rail operator – over concerns that it restricted competition by withholding information, including real-time passenger data, to third-party ticketing platforms.13
As a consequence of skiplagging restrictions, consumers who want flight A-B (e.g. New York to Austin) at a competitive price are forced to also purchase and use flight B-C (Austin to Denver). Here, the tied product is the B-C flight (Austin to Denver) which has no additional financial cost, but which creates no utility. In fact, it creates a considerable utility cost to the customer. This product (B-C) is therefore – for this customer – an economic “bad” with a negative value.14
Absent market power, an airline would be expected to price A-B at more competitive prices and would be unable to enforce a boilerplate skiplagging clause.15 The tie here therefore leverages market power in the A-B route to force unwanted purchases in the A-B-C route, the prospect of which prevents consumers buying cheaper A-B tickets from third party retailers. This helps to maintain the airline’s monopoly pricing in A-B. In these circumstances, the use of market power to impose the tie (and to suppress output of A-B-C tickets) serves to protect and maintain the ability to use that market power to overcharge in the direct A-B market. The restriction therefore acts to prevent arbitrage that would otherwise benefit consumers.
It might be argued that skiplagging restrictions facilitate a kind of price discrimination that improves consumer welfare. Airlines already price-discriminate through their dynamic pricing, selling tickets at different prices in order to efficiently use their capacity. To do so they apply restrictions to prevent consumers reselling their tickets (intermediaries like Skiplagged can resell but at live rates, not via advance purchases). Such price discrimination helps to expand output by ensuring that flights are operated that otherwise may not be.
However, banning skiplagging does not contribute in the same way. This is because – when the airline sells a seat, it receives whichever ‘price’ it puts on that seat – regardless of whether or not the seat is actually used. Therefore, while the restriction on skiplagging may increase prices and revenues overall, it does not change the incentive to run the flight in question, nor the incentive to run unprofitable flights.
A different pro-competitive argument for this pricing practice might be that absent the skiplagging restriction, an airline would have increased the price of indirect A-B-C routes in order to reduce arbitrage incentives and to protect its ability to charge higher prices on A-B tickets. However, this argument relies on out-of-market efficiencies in the A-C market being weighed against consumer harm that is incurred in the A-B market.16 Such out-of-market efficiencies are not generally considered relevant by the courts (since distortionary cross-subsidization of another market by A-B purchasers is not a recognized goal of antitrust law).
Moreover, it is far from clear that A-C prices would actually increase absent the restriction. Direct A-C prices face little or no constraint from indirect A-B-C prices. This means that withdrawing A-B-C offers would not reduce the constraints on direct A-C prices. Furthermore, indirect A-B-C prices can also be assembled by consumers or intermediaries who would be paying less on the A-B leg even if the B-C price were to increase. This would suggest that the net price on A-B-C tickets may not be reduced.
Finally, the airline might, absent the skiplagging restriction, increase B-C prices to reduce the possibility of arbitrage by customers building their own A-B-C routes. This however would be an out-of-market efficiency claim, and not relevant to competition on the B-C route. Moreover, it is unclear that there would be any benefit to consumers in terms of lower B-C prices.17 Ultimately, the airline’s ability and incentive to increase direct B-C prices will depend on direct B-C competition and demand. The case for an efficiency defense therefore appears a weak one.
Given that the U.S. airline industry generated $179 billion in revenue from passenger fares, and around 15% to 17% of flights are affected by HCT, we estimate that the volume of affected commerce was around $30 billion in 2023. HCT saves an average of around $24 per ticket,18 and U.S. airlines carried 658 million passengers in 2021.19 Therefore if consumers skiplagged where feasible they would make savings in the order of $2 billion per year.20
1 www.aa.com/i18n/customer-service/support/conditions-of-carriage.jsp (accessed September 16, 2024).
2 See, e.g., United Airlines et al. v. Zaman, Complaint, No. 14-09214 (N.D. Ill.), ¶13. See also www.npr.org/2023/08/23/1194998452/skiplagging-airfare-flying-skiplagged-american-airlines
3 https://skiplagged.com/about
4 American Airlines v. Skiplagged, Complaint, No. 23-00860 (N.D. Tex.), available at www.documentcloud.org/documents/23923387-american-airlines-v-skiplagged
5 Luttmann, A. and Gaggero, A., “Discount opportunities in hub-and-spoke networks: The determinants of hidden-city ticketing”, Munich Personal RePEc Archive, 26 July 2022, available at mpra.ub.uni-muenchen.de/113960/1/MPRA_paper_113960.pdf (hereinafter, “The determinants of HCT”), p. 30.
6 “Aviation Competition – Restricting Airline Ticketing Rules Unlikely to Help Consumers”, United States General Accounting Office, July 2001, available at www.gao.gov/assets/gao-01-831.pdf, p. 33.
7 The determinants of HCT, p. 17.
8 The determinants of HCT, p. 1. The study notes that “Relative to Sun Country, the omitted airline fixed effect in the regressions, HCT opportunities are approximately 21% more likely on American, 37% more likely on Delta, and 25% more likely on United” (pp. 23-24).
9 The determinants of HCT, p. 6.
10 The determinants of HCT, p. 30.
11 The determinants of HCT, p. 4.
12 www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2023/28_06_2023_DB_Mobilitaet.html
13 ec.europa.eu/commission/presscorner/detail/en/ip_23_3455
14 In economics these are referred to as “bads” rather than “goods”. An example is horse manure which turned from an economic good to an economic bad. ideas.repec.org/a/wej/wldecn/34.html
15 These competitive prices would nevertheless still be dynamically set in line with yield management principles, rather than simply reflecting marginal cost plus a return.
16 For example, it requires using the possibility of cheaper prices in the AC or BC market as a justification for increased prices in the AB market.
17 Analyzing those effects would require looking at the market conditions in the distinct BC market, and BC pricing is likely to be driven by demand and supply for direct travelers rather than the small proportion that travel indirectly.
18 The determinants of HCT, pp. 12-13.
19 www.bts.gov/newsroom/full-year-2022-us-airline-traffic-data
20 We calculated 15% of 658 million and multiplied this by $24.
Emily joined Fideres in December 2020. She holds a degree in Music from Durham University and went on to complete her Law degree at The College of Law in London. After graduating, Emily gained litigation support experience at an international law firm in London. Prior to joining Fideres, she worked as a paralegal at a law firm in New York City, where she was involved in a securities litigation and assisted with a number of blog posts covering trends in the credit markets.
Arbitration clauses limit consumers’ ability to pursue collective actions.
Insights from the Stanford clearinghouse.
No – It was the picture of fairness.
London: +44 20 3397 5160
New York: +1 646 992 8510
Rome: +39 06 8587 0405
Frankfurt: +49 61 7491 63000
Johannesburg: +27 11 568 9611
Madrid: +34 919 494 330