The effects of major U.S. domestic airline code sharing and profit sharing rule
| Date | 01 September 2017 |
| DOI | http://doi.org/10.1111/jems.12202 |
| Author | Caixia Shen |
| Published date | 01 September 2017 |
Received: 27 January 2015 Revised: 8 December 2016 Accepted: 15 December 2016
DOI: 10.1111/jems.12202
ORIGINAL ARTICLE
The effects of major U.S. domestic airline code sharing
and profit sharing rule
Caixia Shen
School of International Business Admin-
istration, China Industrial Develop-
ment Institute, Shanghai Universityof
Finance and Economics, Shanghai, China
(Email: shencaixia@gmail.com)
Abstract
This paper presents a structural model of code sharing among major U.S. domestic
airlines and estimates a profit-sharing rule. The profit-sharing rule between partner
firms in code sharing is estimated at 0.92, which suggests that the operating carrier
acquires around 92% of profits from a round-trip, and the marketing carrier retains
8% as a commission fee. Meanwhile, the economies of code sharing reduces marginal
cost, and firms are able to price at higher markups. This implies that demand increases
and consumers have larger surplus if code sharing creates new products.
1INTRODUCTION
My primary goal is to measure the extent to which code sharing agreements between major U.S. domestic airlines affected
consumer surplus, firms’ costs, and social welfare. Code sharing is an agreement between two (or three, in one case) carriers,
in which one carrier (the operating carrier) allows another carrier (its code-share partner, the marketing carrier) to market and
sell seats on some of the operating carrier’s flights under the marketing carrier’s reservation code.1When a proposal of code
sharing between two major airlines is initiated, a debate often ensues between the policy-makers (U.S. Department of Justice)
and carriers. The U.S. Department of Justice hesitates to approve such proposals due to concerns that these agreements may
reduce competition and hence induce a loss of consumer welfare, given the factt hat potentialcode shar ing partners already have
high market shares. However, carriers have claimed that passengers benefit from code sharing because it provides more product
choices and destinations. This paper lays out the first step and explicitly models these code sharing activities and quantifies its
effects on consumers, carriers, and social welfare.
The model considers a discrete choice setting for consumer demand and an oligopolistic airline supply side. To incorporate
code sharing into the model, I offer an approach that involvesspecifying the supply side as multi-product firms maximizing profits
by choosing a set of prices when they share profits for code-shared products according to a certain rule, which is parameterized
by 𝜆. Moreover, mymodel also takes into account how the economies of code sharing affect fir ms’ marginalcosts, and quantifies
consumers’ direct (non-pecuniary) preference for code-shared flights.
The estimation results suggest that the operating carrier extracts around 92% of profits on average for a code-shared product,
which means that the ticketing carrier retains 8% as a commission fee. The identification of this profit-sharing rule comes from
the variation between code sharing tickets and non-code sharing tickets. In the same market, where I observe both code sharing
and non-code sharing tickets, the data allow my model to estimate 𝜆.
The results show that code-shared products are priced cheaper by the marketing firms compared to prices of products owned
by the marketing firms. Also, code sharing has a strong effect on cost saving so that partner firms can price code-shared products
at higher markups, although consumers have slightly higher elasticities for these products. The sources of cost reduction may
This is a revised versiono f Chapter 1 in mydisser tation written at Boston University.I thank Marc Rysman and Jordi Jaumandreu for advice and encouragement.
I also thank the co-editor and two anonymous referees. This paper has improvedconsiderably due to their comments. I also thank Itai Ater, Tobias Klein,Josh
Lustig, Michael Manove, Naoaki Minamihashi, and Sambuddha Ghosh for comments. Financial support by National Natural Science Foundation of China
(Grant No. 71403161) is acknowledged.Er rors remain myresponsibility.
J Econ Manage Strat. 2017;26:590–609. © 2017 WileyPeriodicals, Inc. 590wileyonlinelibrary.com/journal/jems
SHEN 591
Type 1: vertical
A
operating by United Airlines
−−−−−−−−−−C
operating by U S Airways Airlines
−−−−−−−−−−−−
marketing by U S Airways Airlines
B
Type 2: horizontal
A
operating by United Airlines
−−−−−−−−−−−−−−−−−−−−−−−
marketing by U S Airways Airlines
B
FIGURE 1 The two types of code sharing flights
come from the increase in passengers and the economies of code sharing, which allows airlines to specialize along different
routes. Brand loyalty and an improvement in marketingco-operation af ter code sharing are factorst hat have a positive effect on
filling the capacities of airplanes.
The welfare consequence of code sharing is investigated in two selected markets. I separate the counterfactual analysis into
two cases. In the first case, code sharing does not create new products. The counterfactual analysis simulates a new set of
prices without code sharing, keeping the number of products unchanged. In both markets, I find a small amount of reduction
in consumer surplus due to the negative direct utility effect. The second case is when code sharing induces the entry of new
products. In this case, consumer surplus and firms’ profits are calculated taking into account the entry of the new (code-shared)
products. The results show that code sharing generates large welfare gain from the new products; in particular, both consumer
surplus and firms’ profits increase significantly. In the market with a bigger market size, the total welfare gain is larger. If a
particular code sharing agreement follows in between these two extreme scenarios, one can expect consumer may benefit. This
result shows some evidence to justify the current approvals of code sharing agreement by the Department of Justice.
The remainder of the paper is organized as follows. Section 2 explains code sharing agreements in detail; Section 3 reviews
the related literature; Section 4 outlines the model; Section 5 describes the data; Section 6 proposes the empirical strategy;
Section 7 reports and discusses the estimation results and the counterfactual analysis; Section 8 presents the conclusions.
2CODE SHARING
In practice, there are two types of code sharing: vertical and horizontal. Figure 1 illustrates one example of these two types.2
Vertical code sharing happens when each partner flies part of the flight route on which they code-share. Horizontal code sharing
only refers to the situation when the operating carrier operates the entire route while its partner (the marketing carrier) sells the
tickets. It is important to note that the same flight could be listed twice with different codes in the computer reservation system
(CRS). For example, UA 54 is the same flight as US 6001, which is from Kona, Hawaii to San Francisco, operated by United
Airlines but also code-shared with US Airways Airlines. In this case, there are two ticketing carriers for selling the same seats
on a flight.
For vertical code sharing, the code-shared flight is a product that each partner operates partially. It may or may not exist
before code sharing and could be sold by one carrier or both carriers. However, vertical code sharing does vertically integrate
two carriers in the way that only one carrier markets the tickets as a whole product. For horizontal code sharing, it appears that
one airline sells its partner’s tickets; however, the marketing and operating roles differ across markets.
3LITERATURE REVIEW
Over the past two decades, increasing attention has been paid to code sharing agreements. The related literature begins with
theoretical papers on international code sharing alliances. These models are Park (1997) and Brueckner (2001), both of which
To continue reading
Request your trial