On sources of market power in the airline industry: Panel data evidence from the US airports

Transportation Research Part A 59 (2014) 288–305

Contents lists available at ScienceDirect

Transportation Research Part A journal homepage: www.elsevier.com/locate/tra

On sources of market power in the airline industry: Panel data evidence from the US airports Volodymyr Bilotkach a,⇑, Paulos Ashebir Lakew b a b

Newcastle University Business School, 5 Barrack Road, Newcastle-upon-Tyne NE1 4SE, United Kingdom University of California, Irvine, Institute of Transportation Studies, 4191 AIRB, Irvine, CA 92697, United States

a r t i c l e

i n f o

Article history: Received 19 March 2013 Received in revised form 25 September 2013 Accepted 24 November 2013

Keywords: Market power Airline industry Airport concentration

a b s t r a c t A firm can obtain market power through its dominant position on the product market, or via control of a key resource. In particular, it has been argued that airport dominance is a more important source of market power in the US airline industry than route dominance. We examine this contention by analyzing a seventeen-year panel of airport-level prices in the United States. We demonstrate that even though on average airport-level concentration appears to be the strongest source of market power, concentration on routes originating at an airport is the strongest predictor of price levels for the sub-set of large and medium hub airports. There is little evidence that either airport or route dominance significantly affect prices in the sub-sample of medium and small hub airports. There is also little evidence that an airport’s dominant carrier exerts market power beyond the level predicted by the airport or route dominance. Our results imply that consumer welfare losses due to airline consolidation can be concentrated in smaller communities, and related to changes in airport-level concentration. We provide a simple evaluation of the possible effects of two recent and one projected mergers (Delta-Northwest, United-Continental, and American-US Airways) in light of this finding, and suggest that the former consolidation event can potentially lead to non-trivial consumer welfare losses to travelers in over 30 small communities. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction This study takes a 30,000-feet view at the issue of sources of market power in the US airline industry. We take a fairly straightforward approach. Consider a passenger contemplating a trip that originates at a local airport. The question we ask is: will this passenger pay a higher price, other things equal, if his local airport is served by a few airlines, or if there is little competition on the individual markets originating at this airport? To give the reader a more concrete example of our question: will trips originating at an airport served by only two airlines that compete on each market originating from this gateway be priced higher than tickets for flights, originating at an airport served by four airlines, where each carrier is a monopolist on the respective routes it serves? Additionally, we inquire whether ticket prices will be higher if one of the carriers serving the airport has a dominant position at this gateway, meaning that it carries most of the passengers out of and into the airport. In other words, we examine the role airport dominance, airport concentration, and route concentration play in determining the prices for airline tickets originating at a particular airport. While the commonly held view is that airfares are determined solely by the competition at the individual market level, we can suggest the following rationales for the airport-level concentration to affect average fares for trips originating at an airport, irrespective of whether the airlines compete head-to-head on individual non-stop routes originating at the airport. ⇑ Corresponding author. Tel.: +441912081648. E-mail addresses: [email protected] (V. Bilotkach), [email protected] (P.A. Lakew). 0965-8564/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tra.2013.11.011

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

289

First, airlines operating hub-and-spoke networks will inevitably compete on one-stop routes that originate at the airport, flying passengers to the same destinations via different hubs. Second, airlines enjoying dominant position at an airport can influence airport decision making, including blocking further entry (Berry, 1990). Third, airlines’ entry into an airport can be construed by incumbent carriers as a competitive threat, affecting fares on routes not directly served by the entrant (Goolsbee and Syverson, 2008). Finally, we can expect airlines at less concentrated airports to compete more aggressively for frequent fliers residing in the airport’s catchment area This expectation is drawn from Borenstein’s (1989) original suggestion that frequent flier programs could be a source behind the observed hub premium. While research on pricing in the US airline industry has been extensive, a simple question of the interplay of airport and route dominance has not been examined across the entire industry and over time. Most studies to date have either focused on a limited number of markets, or examined the issue of sources of market power over relatively short time periods. As a result, for instance, little is known about the sources of market power at small airports. Further, a simple question of the relationship between price and airport concentration has been overlooked in the literature, as researchers have focused on estimating market power of the dominant airline at an airport. Our study starts filling these gaps. We have constructed a seventeen-year panel of airport-level average airfares for domestic trips originating at all US airports, using the data collected by the US Department of Transportation (DOT). We have then used a different DOT dataset to construct measures of concentration, at both the airport-level and for the routes with non-stop services originating at the airport. Applying conventional panel data-analysis techniques, and accounting for endogeneity of the key market concentration variables, we determine that, on average, airport concentration stands out as the most important driver of airport-level fares. The size of the effect is however not impressive: a 10% increase in airport-level HHI only raises the average fares by 1– 1.8%. Further analysis reveals that the airport concentration–price relationship is primarily driven by small airports. Data analysis for large and medium hubs (airports that have traditionally been the focus of hub dominance premium studies) reveals route dominance as the primary source of market power: coefficient estimates suggest that a 10% increase in mean HHI for non-stop routes originating from such airports increases average airfares by around 7%. We also find sporadic evidence suggesting that increase in the market share of the largest airline in the airport may increase airfares. Our research has implications, in particular, for analyzing effects of mergers in the airline industry. Recent debates over this important issue have focused on addressing the question of the impact of mergers on airfares via changes in the routelevel competition (Brueckner et al., 2013). We show that analyzing effects of airline consolidation on airport-level concentration is also important. Specifically, our analysis shows that the Delta-Northwest merger is projected to substantially increase concentration at about three times as many airports as the Continental-United merger. The projected American-US Airways merger is projected to increase airport-level concentration in a number of otherwise not concentrated medium and large hub airports. We further demonstrate that the Delta-Northwest merger can potentially lead to about a quarter of a billion dollars per year in terms of consumer welfare losses through increased airport-level concentration at around thirty small airports. More generally, we suggest that gains and losses from airline consolidation may not be distributed evenly across the US airline industry; and in particular, losses from consolidation could be disproportionately borne by smaller communities. The rest of the study is organized as follows. The next section reviews the relevant literature, followed by the description of the data and discussion of the analysis methodology. After this, data analysis results are presented and discussed, paying specific attention to the implications of our exercise for analysis of the airline mergers. The last section of the paper offers some concluding remarks. 2. Relevant literature Looking at the rather extensive literature on the sources of market power in the US airline industry, two stylized facts stand out. First, low-cost carriers (LCC’s), in particular Southwest Airlines, push average fares down on the routes they enter. Second, airport dominance appears to play an important role as a source of market power; perhaps more important than the route dominance. These facts reflect the focus of the empirical studies on LCC’s and – to a larger extent – hub operators, which typically dominate the airports they serve. This focus is understandable, as emergence of hub-and-spoke networks and explosive growth of low-cost carriers have been the two most important innovations in the airline industry since deregulation. The path along which the literature developed has, however, led to the following important gaps, which our study intends to start filling. First, airport-concentration–price relationship has received substantially less attention than the dominant-carrier’s-market-share-price correlation.1 Second, smaller airports, which are as likely to be dominated by a single carrier, received little coverage. Third, aggregate-level studies examining the issue of sources of market power over longer time periods are scarce: a recent paper (Borenstein, 2011) offered such an analysis, focusing on airline-market-level fares, and demonstrated that the magnitude of hub premium has declined over the recent years. Finally, studies of the effects of airline mergers tend to assess average system-wide effects of the consolidation events, paying relatively little regard to potentially important differences between groups of routes or origin airports. 1 Van Dender (2007) offered a direct evaluation of the relationship between airport-level fares and airline concentration. The study used a simultaneous equation system approach to present an analysis of determinants of airport-level prices, delays, frequency of service, and operating revenues at 55 large airports in the US. Van Dender did not find any significant relationship between airport-level concentration and prices. Compared to this study, our analysis provides a more comprehensive coverage of the US airline industry. Using a longer panel, we analyze price effects of the largest carriers’ market share and the competition on routes originating at an airport.

290

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

While addressing the implications of the hub-and-spoke networks, early post-deregulation studies of the airline industry focused on testing the market contestability hypothesis, which asserted that potential competition could put pressure on the exercise of market power by the incumbent carriers. Notable studies include Bailey and Panzar (1981) and Hurdle et al. (1989), which rejected the hypothesis that all city-pair markets are equally contestable, and found that the number of potential entrants had a significant negative effect on yields. Following establishment of hub-and-spoke networks, seemingly higher fares at concentrated hub airports turned researchers’ attention to the hub premium. The General Accounting Office’s (GAO) widely cited 1990 study was the first to quantify the hub premium when it found that yields for trips originating at dominated hub airports were 27.2% higher than yields at airports that were not concentrated in 1988.2 However, the GAO’s study was also criticized for implicitly assuming that trips taken from these two groups of airports were identical, not taking into account route distance, the number of plane changes, passenger mix, carrier identity, unit-cost differences, and frequent flier programs (FFPs). Similar analyses that controlled for some of these factors were carried out and found more modest levels for the hub premium.3 Borenstein (1989) is regarded as the seminal econometric study examining the airport dominance premium. Airport dominance and route dominance were found to significantly affect the fares in markets where a carrier is dominant at the originating hub airport. In a follow-up study, Borenstein (1991) showed that an airline with a dominant position at an airport has a larger share of the overall originating traffic, and thereby also has a larger share of any market originating at the dominated hub.4 Subsequent research provided new evidence on the hub reliability premium issue: Evans and Kessides (1993) concluded that airport dominance is a far more important source of market power than route dominance. Lee and Luengo-Prado (2005) found that the hub effect increases as the share of premium class passengers rises; the presence of a LCC reduces premium fares by as much as 6%, while reducing economy-class fares by as much as 14%. Overall, they suggested that much of the observed hub premium could be explained by the passenger mix.5 Bilotkach (2007) demonstrated that dominant carriers command the corresponding premiums on the international routes. Two studies documenting the impact of low-cost carriers are Morrison (2001) and Brueckner et al. (2013). Morrison (2001) showed that actual and adjacent competition from this airline has saved passengers $12.9 billion in 1998. Brueckner et al. (2013) demonstrated that competition from the legacy carriers does not affect airfares, while competition from the low-cost airlines does. Borenstein (1990) and Kim and Singal (1993) examined the price and market power effects of the 1980s wave of airline consolidations. Clougherty (2002) suggested that US airline mergers improve the international competitiveness of US carriers. Richard (2003) looked at the welfare effects of a hypothetical merger between American Airlines and United Airlines for Chicago-originating routes, focusing on the carriers’ choice of fares and frequency. The consequences of eliminating potential competition were explored by Kwoka and Shumilkina (2010) while Bilotkach (2011) looked at the effects of changes in multimarket contact on the intensity of competition, effectively evaluating coordinated effects of a consolidation event. Most recently, Luo (2013) addressed the Delta-Northwest merger, suggesting that this event resulted in a small increase in airfares at the market level (looking at both non-stop and connecting routes). Our study does not disagree with Luo’s findings; however, we specifically point to those markets where the Delta-Northwest merger could have resulted in increased airfares. We also suggest that any adverse effects of an airline merger could indeed be localized to smaller communities.6 3. Data Data for our research exercise come predominantly from Bureau of Transportation Statistics (BTS), a division of the US Department of Transportation (DOT). BTS collects information on airline operations and airfares in the US airline industry. Average airport-level airfares we will use as the dependent variable are computed by the DOT from the 10% sample of actual itineraries.7 We use annual average airfares for seventeen years, from 1993 until 2009, making adjustment for inflation using 1993 as the base year. Thus, our dependent variable represents annual average ticket price, in constant dollars, paid by a pas2 See General Accounting Office, Airline Competition, p. 3. A filtered US DOT Data Base 1A (DB1A) was used to compare 15 dominated hub airports to a control group of 38 non-concentrated airports. 3 In 1990, the US DOT also examined these two airport groups while controlling for route distance. The DOT study found an average hub premium of 18.7% and 8.9% for airports dominated by one and two airlines, respectively. Morrison and Winston (1995) revised the GAO study, controlling for the five factors mentioned above. They found a significantly lower hub premium of 5.2%. In particular, corrections for distance and the number of plane changes reduced the estimated premium by 18.6% points; carrier-specific comparisons shed off another 4.6% points; correcting for FFPs and passenger mix removed each removed yet another 2.5% points off the premium. Dresner and Windle (1992) controlled for stage length and airport-level characteristics when comparing yields on flights that originate from hubs to yields on flights that are destined to hubs. Since market power is more pertinent at origin airport, flights from a hub were expected to render higher yields than those to a hub. Their results showed the existence of a statistically significant 1–2% hub premium. 4 Fruhan (1972) addressed the underlying idea of this effect by showing that an airline with a large share of available route capacity would disproportionately attract more traffic on that route. This route-based capacity advantage was also supported by Bailey et al. (1985) in their review of the deregulated industry. The strong evidence that the authors found for route-based capacity advantage prior to deregulation had diminished by 1981. Ciliberto and Williams (2010) also showed that control of an airport’s resources appears to be an important source of the dominant carriers’ market power. 5 Lederman (2008) found that FFPs account for 25–37% of the hub premium that carriers charge. 6 See Tretheway and Kincaid (2005) for a comprehensive review of the literature on market structure and airfares. 7 The airfares we use are averages provided by the DOT. According to the agency, fares are computed from DB1B dataset, lumping together fares for one-way and roundtrip itineraries originating at an airport, and excluding very low (under $50) or very high (yield exceeding $3) priced itineraries.

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

291

senger departing from an airport. We should note that, given the methodology used by DOT to obtain the data, within-airport variation in airfares can potentially be affected by changes in both price levels and the share of roundtrip itineraries (around 80% of trips in DB1B). However, we have no reason to believe that any changes in the proportion of one-way versus round-trip itineraries out of any airport could be driven by changes in any of the key measures of market structure we employ here. The majority of our independent variables are computed from the DOT’s BTS (Form 41 Traffic) T-100 Segment tables. This tables include information about all commercial airline services departing from US airports. The information is provided monthly at the airline-origin–destination-aircraft type level (e.g., Delta Air Lines Boeing-757 services from Los Angeles International to New York John F. Kennedy airport in January of 2000 are recorded separately from the Boeing-737 services of the same carrier on the same route in the same month) and includes the number of flights performed, seats provided, and passengers carried. We have aggregated the data at the annual level, and merged regional airlines with the respective major carriers.8 After this, we have computed the following variables from T-100 Segment tables, at the airport-year level. Airport level Herfindhal-Hirschmann Index (HHI). HHI is the sum of squared market shares, across all the airlines at an airport. We have computed the index based on airlines’ passenger volumes. For the purpose of this exercise, we re-coded flights by regional carriers as those of the respective major airlines. Highest market share. This variable represents the share of passengers, handled by the largest airline at an airport. This is effectively a one-firm concentration ratio. Route HHI. This variable is computed as the passenger-weighted average HHI on all non-stop routes originating at the airport. This serves as a measure of route dominance. We should however note that this is an imperfect measure of competition between the airlines. In particular, many smaller airports may only offer services to the major carriers’ hub; but some passengers originating at those smaller airports will continue their journeys beyond those hubs. Then, even if we find that all individual routes originating at an airport are monopolized, the airlines might still be competing on one-stop markets going through their hubs. The three variables above will be the key measures of market structure, and the impact of those variables on the mean airport-level price will be the focus of our research. The three measures are naturally correlated with each other. In particular, the correlation coefficient between HHI and highest market share is over 0.9, whereas the correlation between route HHI and the other measures of concentration is around 0.6. We will need to be mindful of these facts when interpreting the estimation results. The following control variables have also been computed from the T-100 tables. Number of destinations served from the airport. This is a simple count of the number of unique airports within the US, served with non-stop scheduled passenger flights from the airport in every given year. In this computation, we only included those destinations served with at least 100 flights per year – roughly two flights per week were thus required for a destination to be counted. Total airport-level passenger volume. This variable will control for possible scale effects. Mean distance. This variable was calculated as the passenger-weighted mean distance of a non-stop flight from the airport. This distance clearly underestimates the total one-way distance in an average passenger’s itinerary. Market shares of individual airlines. Market shares have been computed for thirteen individual major carriers, some of which have not been in existence for the entire duration of our panel. The main purpose of this variable will be to account for possible price changes due to growth of LCC’s (most importantly, Southwest Airlines and JetBlue Airways). See notes to Table 6 for more details. Other control variables we will use include population, average weekly wage, and unemployment rate. These variables are measured at the corresponding Metropolitan Statistical Area (MSA) level, and have been obtained from the US Census Bureau and the US Bureau of Labor Statistics’ Quarterly Census of Employment and Wages (QCEW). Table 1 includes basic descriptive statistics of our variables. We present means and standard deviations for the entire sample, as well as for the three sub-samples we will focus on later in this study. To define sub-samples, we use the Federal Aviation Administration’s (FAA) classification of airports. The FAA classifies airports as primary and non-primary, using 10,000 passenger boardings per year as the cut-off. For our analysis, we will only focus on primary airports, which are further subdivided into non-hub airports, small hubs, medium hubs, and large hubs based on the percentage of total passenger boardings handled by the airport. Specifically, non-hub airports are primary airports which handle less than 0.05% of total passenger traffic on the US market.9 Small hubs handle 0.05–0.25% of all passengers. In order to be classified as a medium hub, the airport needs to handle more than 0.25%, but less than 1% of all passengers. Finally, airports that handle over 1% of all passenger boardings in the US market are classified as large hubs. Note that some airports may change their classification over the years. Overall, there are 442 airports in our dataset, which are classified as primary in at least one year. Given the data analysis methodology we use, lagged endogenous variables as instruments in particular, 399 of those airports are actually included into our regressions. The following facts are apparent from the data. First, an average airport is rather concentrated. At the same time, smaller airports are much more concentrated than larger hubs. In all fairness, large hubs are more concentrated than medium hubs; 8 Some services, in particular on thinner routes, are delegated by the major carriers to regional airlines, typically using smaller jet and/or turboprop aircraft. The original T-100 Segment tables code regional airlines differently from the majors. Details on the procedure used to merge regionals with the majors are available from the authors upon request. 9 To put this into perspective, 0.05% of total US domestic passenger traffic corresponds to about 300,000 passengers per annum.

292

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

Table 1 Descriptive statistics.

Real airfare (year 1993 dollars) Airport HHI Highest market share Route HHI Number of destinations Total passenger volume Mean distance, miles one-way Real wage, constant $/week Population, MSA level Unemployment rate (%)

All airports

Large and medium hubs

Medium and small hubs

Small hubs and non-hub

323.20 (129.68) 0.6134 (0.3146) 0.6775 (0.2775) 0.9257 (0.0819) 33.02 (39.12) 1,783,273 (4,302,732) 411.80 (396.18) 483.56 (102.69) 1,006,101 (2,634,735) 5.43 (2.50)

247.61 (67.17) 0.3137 (0.1920) 0.4446 (0.1942) 0.8502 (0.0569) 91.37 (41.02) 7,520,611 (6,666,942) 880.84 (453.91) 589.67 (108.84) 3,663,681 (4,160,709) 5.00 (1.70)

280.54 (85.35) 0.3481 (0.1836) 0.4347 (0.1989) 0.8930 (0.0604) 47.42 (29.13) 654,037 (334,820) 460.35 (230.31) 499.19 (77.69) 1,235,492 (3,385,086) 5.17 (2.25)

350.56 (139.95) 0.74450 (0.2716) 0.7878 (0.2375) 0.9515 (0.0770) 17.16 (17.66) 60,225 (75,698) 285.21 (310.67) 451.43 (84.41) 238,378 (783,881) 5.62 (2.71)

Note: this table includes mean values, with standard deviations in parentheses. Only data for primary airports (those handling over 10,000 passengers per year) are used in calculations. We use FAA airport classification, as described in this section of the paper. We have a total of 442 unique airports that are classified as primary at least once.

Fig. 1. Dynamics of concentration measures and highest market share.

however, average airport-level HHI for this group, at 0.35, is still half of the same for small hubs and non-hub airports. Airports’ largest carriers handle substantial share of passenger traffic, as evidenced by the average values of the highest market share variable. Note that the largest carriers at smaller airports tend to carry larger shares of passenger traffic as compared to the same for bigger gateways. Further, non-stop markets tend to be highly concentrated, with larger airports exhibiting more route-level competition. An average airport is connected to 33 other US airports with regular scheduled non-stop passenger services, and larger airports understandably exhibit more extensive connectivity to the US aviation network. Last but not least, smaller airports tend to be located in less populous areas (which is hardly surprising in itself) with lower wages and higher rates of unemployment. The purpose of Fig. 1 is to give the reader an idea of the behavior of our main variables over time. Observe that both airports and non-stop markets have exhibited the highest concentration in the 1990s. The dips in both HHIs and the highest market after 2001 can be attributed to a series of high-profile airline bankruptcies and reorganizations following the events of September 11, 2001. Specifically, United, Delta, Northwest, and US Airways spent some time under Chapter 11 bankruptcy protection, which resulted in downsizing and some hub closures.10 At the same time, low-cost carriers, most notably Southwest Airlines, used the temporary weakness of the legacy carriers to encroach on their territory (see Boguslaski et al., 2004, for more details). JetBlue Airways has also been growing rapidly after being founded in 2001 (see Bilotkach et al. (2012) for a discussion of JetBlue’s network development).

10

US Airways dismantled Pittsburgh and Baltimore hubs; Delta pulled out of Dallas-Fort Worth.

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

293

Fig. 2. Average airfares and airport concentration for large and medium hubs.

Figs. 2 and 3 show the average airfares at the large and medium airports in our sample, over the years 1993–2009, with average airport and route concentration (HHI), respectively. The two figures illustrate the airport-level price–concentration relationship, which is central to this study. Our analysis, however, captures the airport-level price effects through the temporal variation of concentration and control variables. 4. Hypotheses and methodology The focus of our study will be on the relationship between airport-level prices, airport concentration, largest carrier’s market share, and concentration on routes originating at the airport. We clearly expect higher concentration and an increase in the dominant airline’s market share to increase the fares. The central question then is: all else equal, which of these three factors is a more important predictor of price level? We will also investigate whether sources of market power at the origin airport level differ across groups of airports. The two econometric problems we will need to tackle are airport-level heterogeneity and potential endogeneity of the three measures of market concentration. We have chosen conventional approaches to address these problems. Specifically, airport-related idiosyncrasies will be captured by the airport fixed-effects model, and an instrumental variable approach will be employed to address the endogeneity concern. Lagged endogenous variables will be used as instruments in all specifications. Among the control variables, airport-level passenger volume is potentially endogenous; lagged passenger volume will be used as an instrument for this variable. Finally, standard errors we report are robust to both heteroscedasticity and autocorrelation – our data structure clearly points to the threat of both unequal error variances across, and correlation of the current error with its past realizations within the cross-sections. Variables used as instruments in two-stage least squares must meet the two textbook requirements of being correlated with the endogenous variable and uncorrelated with the error term. The former requirement is easily met by the lagged endogenous variables11 (so that we can be certain that the instruments we have chosen are not weak). However, we can be rightfully concerned that current price levels could still be correlated with last year’s unobserved shocks which have affected our instruments. To address this concern and demonstrate validity of our instruments, we will apply the standard Sargan–Hansen test in our estimation. Before we proceed, we should note that our choice of methodology drives the interpretation of our coefficient estimates. The use of airport-level fixed effects – a logical choice given the dataset we have – implies that identification of our regression coefficients will be driven by the variation in variables within airports, not across them. For instance, our coefficient on airport-level HHI will tell us how the airport-level price will respond to the increase in airport-level concentration within an airport over time, not whether more concentrated airports tend to be associated with higher average fares at any given point 11

Correlations between endogenous variables and their lags in our dataset is 0.8 or higher.

294

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

Fig. 3. Average airfares and route concentration for large and medium hubs.

in time. The latter is an appropriate interpretation of the regression coefficient in a cross-section. As a side note, using a longer panel, we should be able to get sufficient within-airport variation in the key concentration measures to identify the relationships we are looking for. We will return to this issue when discussing the estimation results. As described in the previous section, the control variables we will be using include number of destinations served with non-stop flights, total passenger volume, mean non-stop distance of a flight originating at an airport, airport-level market shares for the major players in the US airline industry, population, and two socioeconomic measures for the corresponding metropolitan statistical area. In addition to these, we will also use year dummy variables to control for the corresponding heterogeneity. We will start by performing our analysis for the entire sample. Subsequently, we will evaluate whether the concentration–price relationships differ across the airport groups. For this, we will first define three sub-samples of airports, based on their relative size in the US aviation system (as classified by FAA). One sub-sample will include large and medium hubs; the second sub-sample will contain small and medium hub airports; and the third sub-sample will include small hubs and primary non-hub airports. We have chosen not to conduct the analysis separately for large, medium, and small hubs, as each individual category includes rather few airports (37, 48, and 84, respectively).12 Most primary airports are, not surprisingly, classified in the non-hub category. For both the entire sample and each individual sub-sample, we will run 6 specifications, which will differ in terms of the combinations of the key concentration variables employed. We will begin by including only one of the three concentration measures into regression. Then, we will estimate specifications including two such measures (airport-level and route-level HHI), highest market share, and route-level HHI. Finally, a specification that includes all three concentration measures will be estimated. To enable interpretation of regression coefficients as respective elasticities, natural logarithms of all measures of concentration will be used. Other control variables, except for the airline market shares and year dummies, will be also included in logarithmic form. We should note that the correlation between airport-level HHI and largest carrier’s market share is very high in our sample. We can therefore expect that including both variables into the same specification might produce insignificant coefficients, even though both measures of concentration could be important determinants of airport-level prices. We should therefore be cautious when interpreting the corresponding estimation results. At the same time, lower correlation levels between route-level HHI and airport-level concentration measures give us more confidence to suggest that specifications where both these variables are present are able to disentangle the corresponding effects.

12 These represent counting of the airports, which fell into the said categories in at least one year. Due to the nature of the classification criterion used, airports can change their affiliation across the categories over time.

295

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305 Table 2 Results for all airports.

Constant Airport HHI

(1)

(2)

(3)

(4)

(5)

(6)

3.1420 (1.4180) 0.1261 (0.0332) –

3.3185 (1.4304) –

3.0201 (1.3907) –

3.2204 (1.3839) –

–

3.0969 (1.3853) 0.1063 (0.0477) –

3.0568 (1.3876) 0.1325 (0.0527) 0.0365 (0.0438) 0.2897 (0.4306) 0.0565 (0.0177) 0.0716 (0.0351) 0.0007 (0.0474) 0.1691 (0.1232) 0.2230 (0.1197) 0.2404 (0.0594) 0.7670 0.038 (0.8448)

Mean route HHI

–

0.0968 (0.0296) –

Number of destinations

0.0564 (0.0180) 0.0719 (0.0349) 0.0051 (0.0457) 0.1708 (0.1226) 0.2141 (0.1218) 0.2479 (0.0544) 0.7676 0.058 (0.8089)

0.0473 (0.0182) 0.0803 (0.0339) 0.0019 (0.0460) 0.1749 (0.1220) 0.2052 (0.1223) 0.2389 (0.0543) 0.7675 0.039 (0.8434)

Highest market share

Total passenger volume Mean distance Real wage Population Unemployment rate Adjusted R-squared Sargan–Hansen (p-value)



0.5763 (0.3230) 0.0413 (0.0195) 0.0823 (0.0324) 0.0084 (0.0324) 0.1952 (0.1211) 0.2244 (0.1176) 0.2220 (0.0583) 0.7650 0.014 (0.9046)

0.2527 (0.4132) 0.0561 (0.0178) 0.0725 (0.0350) 0.0012 (0.0473) 0.1680 (0.1229) 0.2207 (0.1196) 0.2389 (0.0598) 0.7673 0.038 (0.8448)

0.0691 (0.0449) 0.3515 (0.4191) 0.0481 (0.0184) 0.0797 (0.0334) 0.0031 (0.0477) 0.1713 (0.1225) 0.2160 (0.1193) 0.2287 (0.0591) 0.7668 0.890 (0.3454)

Notes: 1. Method used – airport-level fixed effects two-stage GLS. One-year lags are used as instruments for HHI, highest market share, route HHI, and passenger volume. 2. All specifications use natural logarithms of all independent variables reported here, except for constant. 3. Standard errors, robust to heteroscedasticity across and autocorrelation within cross-sections, are reported in parentheses. 4. Airport-level airline market shares and year dummy variables are included in all specifications, but not reported. 5. Included observations: 4771. 6. Significance: 10%; 5%.

Research hypotheses with respect to our key variables are quite clear – higher concentration is expected to lead to higher average fares. Indeed, our aim is to see which of the measures, if any, are more robust predictors of airport-level airfares, other things equal. With respect to control variables, we expect higher total passenger volume to lower airfares, presumably due to scale effects. Higher real weekly wages and population will be expected to lead to higher fares, whereas higher unemployment rate in the airport’s MSA should lower the level of fares going out of the airport. Population and the two socioeconomic variables can be thought of as demand shifters. Higher average non-stop flight distance should increase average fares. A better connected airport will potentially attract more business traffic (Bel and Fageda, 2008), leading to higher fares. Increases in market shares of the low cost carriers, such as Southwest Airlines or JetBlue Airways) will be expected to lower average airfares. 5. Estimation results Results of our data analysis exercise are presented in the following tables. Table 2 reports results for the entire sample, or rather population of airports. Tables 3–5 present estimation results for the three groups of airports we identified above. Results for large and medium hub airports are shown in Table 3, while Tables 4 and 5 depict results for medium and small hubs; and small hubs and non-hub primary airports, respectively. Finally, Table 6 reports results for the airline market share variables – these come from selected specifications from Tables 2–5.13 Every specification is accompanied by the Sargan-Hansen statistic to address potential concerns that our chosen instruments may be correlated with the error term. The null hypothesis is not rejected in either specification, suggesting that our instruments are indeed valid. Results in Table 2 show the airport-level concentration index as the most robust predictor of average airport-level airfares. Numerically, a 10% increase in airport-level HHI leads to 1.05–1.3% increase in average airfares for flights departing from an airport; a 1 standard deviation increase in this measure of concentration is associated with about a 5% increase in average airfares. There is also some evidence to suggest that the largest carrier’s market share at an airport could be associated with higher average prices paid by the traveling public. Here, a 10% increase in the dominant airline’s market share is associated with about 0.97% increase in average fare (from specification 2 in Table 2). Comparing results for different groupings of airports, we notice substantial differences across them. Results for large and medium hub airports suggest that competition on the routes going into and out of those airports is the most robust driver of 13

The corresponding coefficients in other specifications are similar to those reported in Table 6.

296

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

Table 3 Results for large and medium hub airports.

Constant Airport HHI

(1)

(2)

(3)

(4)

(5)

(6)

2.7789 (2.6760) 0.1201 (0.0425) –

2.1283 (2.7417) –

3.0567 (2.3188) –

3.0324 (2.5025) –

–

3.3033 (2.5095) 0.0594 (0.0447) –

3.3446 (2.3863) 0.0691 (0.1035) 0.0108 (0.0928) 0.6517 (0.3161) 0.1981 (0.0567) 0.2118 (0.0569) 0.5283 (0.1118) 0.2438 (0.1569) 0.2118 (0.1392) 0.1174 (0.0497) 0.9078 0.012 (0.9115)

Mean route HHI

–

0.1027 (0.0438) –

Number of destinations

0.2264 (0.0514) 0.2547 (0.0549) 0.5230 (0.1127) 0.2519 (0.1506) 0.2867 (0.1454) 0.1500 (0.0568) 0.8971 0.016 (0.8979)

0.2176 (0.0513) 0.2403 (0.0603) 0.5214 (0.1147) 0.2382 (0.1559) 0.3084 (0.1432) 0.1488 (0.0547) 0.8959 0.015 (0.9011)

Highest market share

Total passenger volume Mean distance Real wage Population Unemployment rate Adjusted R-squared Sargan–Hansen (p-value)



0.8123 (0.3147) 0.1782 (0.0570) 0.1865 (0.0640) 0.5267 (0.1083) 0.2102 (0.1649) 0.1841 (0.1369) 0.1142 (0.0495) 0.8920 0.010 (0.9192)



0.6502 (0.3188) 0.1975 (0.0564) 0.2115 (0.0572) 0.5276 (0.1113) 0.2628 (0.1583) 0.2140 (0.1439) 0.1172 (0.0545) 0.8971 0.012 (0.9115)

0.0400 (0.0398) 0.7127 (0.3064) 0.1905 (0.0567) 0.1992 (0.0616) 0.5286 (0.1119) 0.2574 (0.1620) 0.2161 (0.1395) 0.1139 (0.0514) 0.8961 0.034 (0.8516)

Notes: 1. Method used – airport-level fixed effects two-stage GLS. One-year lags are used as instruments for HHI, highest market share, route HHI, and passenger volume. 2. Reported results are for sub-sample of airports handling more than 0.25% of all passenger boardings within the US in a given year. 3. All specifications use natural logarithms of all independent variables reported here, except for constant. 4. Standard errors, robust to heteroscedasticity across and autocorrelation within cross-sections, are reported in parentheses. 5. Airport-level airline market shares and year dummy variables are included in all specifications, but not reported. 6. Included observations: 1029. 7. Significance: 10%; 5%.

airport-level airfares. Note also that the statistical significance of both airport HHI and largest carrier’s market share vanishes when we include mean route HHI into our specifications. Numerically, our estimates suggest that a single standard deviation increase in mean route HHI, holding other things constant, will be associated with 3.8–4.6% increase in mean airfare for flights originating at a given airport (depending on which of the four coefficients from Table 3 we will employ). Results for the group of medium and small hub airports suggest that changes in concentration at either the airport or the route level do not appear to affect airfares in this group. One of the specifications in Table 4 produces the expected price effect of changes in the largest carrier’s market share. However, this effect does not appear robust to including other concentration measures. The magnitude of the effect is also not too impressive – a 1% increase in airport level airfares would require about a 15% increase in the share of the largest carrier at the airport. Overall, small and medium hubs tend to be the least concentrated airports – we can thus suggest that, with this group of airports, we are probably in the territory where airlines price quite competitively, and within-airport changes in concentration measures do not therefore have much effect on prices. At the same time, goodness of fit suggests our specifications explain most of the variation in prices across and within these airports. Results for the group of small hub and non-hub airports largely replicate the outcome observed for the entire sample, in terms of statistical significance and magnitude of the coefficients on key variables. Therefore, it appears that the results for the entire sample are indeed driven by this group of airports. To sum up results reported in Tables 2–5, we can say the following. While variation in airport-level HHI is the strongest predictor of changes in airport-level airfares, this result is predominantly driven by the smaller airports. For the large and medium hubs, route-level competition is the most robust predictor of average price. There are no surprises present in the coefficients for control variables. It is true that statistical significance of most variables is patchy, and coefficient magnitudes are not stable across sub-samples. However, where variables are significant, the corresponding coefficients show the expected signs. Specifically, passenger volume exhibits evidence consistent with the scale effects (this result is also present in Van Dender, 2007) and greater flight distance is associated with higher fares, even though we have indicated that distance of a non-stop flight underestimates the distance of an itinerary flown by a passenger commencing his/her trip at a given airport. Average fares increase as more destinations are served from an airport with nonstop flights. Demographic and socioeconomic controls also show the expected behavior. There are no surprises in the effects of the airline market shares either. Southwest Airlines’ market share shows the largest downward pressure on airfares, with some of the other low cost carriers (JetBlue and Frontier) also producing price reduction in some of the samples we analyzed.

297

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305 Table 4 Results for medium and small hub airports. (1)

(2)

(3)

(4)

(5)

(6)

1.0187 (2.1885) –

0.9720 (2.1559) – –

Mean route HHI

–

0.0648 (0.0301) –

0.8035 (2.2128) 0.0319 (0.0559) –

0.8297 (2.2128) –

Highest market share

1.0049 (2.1869) 0.0561 (0.0448) –

Number of destinations

0.0687 (0.0329) 0.3990 (0.0528) 0.3773 (0.0764) 0.5495 (0.1964) 0.4813 (0.1527) 0.2592 (0.0499) 0.9095 0.084 (0.7712)

0.0723 (0.0331) 0.4079 (0.0528) 0.3791 (0.0753) 0.5449 (0.1929) 0.4905 (0.1553) 0.2539 (0.0495) 0.9103 0.083 (0.7732)

0.5179 (0.5144) 0.0577 (0.0366) 0.3823 (0.0495) 0.3842 (0.0782) 0.5721 (0.2004) 0.4475 (0.1461) 0.2353 (0.0380) 0.9086 0.059 (0.8069)

0.4523 (0.5793) 0.0630 (0.0372) 0.3838 (0.0483) 0.3826 (0.0777) 0.5353 (0.1978) 0.4552 (0.1533) 0.2389 (0.0369) 0.9098 0.064 (0.8011)

0.8359 (2.2203) 0.0217 (0.0775) 0.0591 (0.0476) 0.4161 (0.5794) 0.0663 (0.0369) 0.3932 (0.0510) 0.3833 (0.0763) 0.5333 (0.1951) 0.4654 (0.1562) 0.2356 (0.0359) 0.9104 0.159 (0.6899)

Constant Airport HHI

Total passenger volume Mean distance Real wage Population Unemployment rate Adjusted R-squared Sargan–Hansen (p-value)

0.0453 (0.0367) 0.4062 (0.5695) 0.0667 (0.0367) 0.3915 (0.0492) 0.3839 (0.0768) 0.5323 (0.1951) 0.4646 (0.1550) 0.2373 (0.0376) 0.9105 0.065 (0.7979)

Notes: 1. Method used – airport-level fixed effects two-stage GLS. One-year lags are used as instruments for HHI, highest market share, route HHI, and passenger volume. 2. Reported results are for sub-sample of airports handling between 0.05% and 0.25% of all passenger boardings within the US in a given year. 3. All specifications use natural logarithms of all independent variables reported here, except for constant. 4. Standard errors, robust to heteroscedasticity across and autocorrelation within cross-sections, are reported in parentheses. 5. Airport-level airline market shares and year dummy variables are included in all specifications, but not reported. 6. Included observations: 1457. 7. Significance: 10%; 5%.

Of the legacy carriers, American Airlines, Northwest Airlines, and US Airways appear to produce price increases as they expand their airport presence, other things equal. In summary, our results demonstrate the following. First, effects of airport-level concentration on airfares for trips originating at an airport depend on the airport size. Specifically, the relationship is the strongest in smaller airports. There is also some evidence suggesting airport concentration might affect fares in the largest airports. In mid-sized airports, however, concentration does not appear to affect average fares for trips originating at those gateways. Second, route-level concentration appears to be a robust predictor of airport-level fares at the larger airports. Third, the pure airport dominance effect (as proxied by the airport-level market share of the largest carrier) does not appear to be robust to including airport-level and/or route-level concentration measures into specifications. We should however be mindful here of significant correlation between the largest carrier’s market share and airport-level HHI, which could have the potential to inflate the standard errors on the corresponding coefficients. At the same time, the fact that the significance of the highest market share variable does not survive inclusion of the route HHI (correlation between these two measures is not modest, but not worrisomely strong) suggests that the highest market share probably does not affect prices in a robust fashion. Coming back to our question of whether airfares will be higher at airports served by a few airlines competing on each market, or at the gateways served by many carriers, each being a monopolist on the respective route, it appears that the answer will also be different for large and small airports. At large airports, new entry (or reduced airport-level concentration) will only bring down the airfares if it also increases competition on individual routes. That is, only entry at those routes where the entrant competes with the incumbent will affect airport-level fares. In smaller airports, however, new entry will bring average fares down regardless of whether the new carrier comes in with new services, or competing with the incumbents. At the same time, only large-scale entry can appreciably change average route-level concentration in a large hub, whereas relatively small-scale entry might be sufficient to alter airport-level concentration in a non-hub airport. There are some potentially important issues that our analysis does not address. Most importantly, when it comes to measuring competition on routes originating at an airport, we only look at non-stop markets. It is however a well-known fact that airlines operating hub-and-spoke networks may compete for passengers on one-stop markets. Then, even though non-stop routes might not be competitive, airlines will compete for passengers on one-stop markets. Further, the airport-concentration–price relationship we observe might work through competition on one-stop markets. Examining this relationship can be an interesting topic for a future project. Also, a quick look at Fig. 1 shows that airline markets are rather concentrated at the non-stop level. On top of this, within-airport variation in the route-level concentration measure is not

298

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

Table 5 Results for small hub and non-hub airports. (1)

(2)

(3)

(4)

(5)

(6)

3.9742 (1.4904) –

4.0040 (1.4403) – –

Mean route HHI

–

0.0899 (0.0335) –

3.6817 (1.4903) 0.1259 (0.0638) –

3.9397 (1.4650) –

Highest market share

3.6810 (1.5061) 0.1256 (0.0392) –

Number of destinations

0.0443 (0.0183) 0.0568 (0.0373) 0.0168 (0.0450) 0.2166 (0.1508) 0.1566 (0.1437) 0.2603 (0.0633) 0.7392 0.068 (0.7913)

0.0347 (0.0186) 0.0670 (0.0355) 0.0210 (0.0453) 0.2283 (0.1490) 0.1354 (0.1425) 0.2499 (0.0631) 0.7395 0.394 (0.5300)

0.3990 (0.4550) 0.0287 (0.0210) 0.0697 (0.0334) 0.0296 (0.0455) 0.2147 (0.1498) 0.1454 (0.1498) 0.2444 (0.0685) 0.7378 0.032 (0.8567)

0.0034 (0.6103) 0.0442 (0.0183) 0.0568 (0.0374) 0.0168 (0.0472) 0.2167 (0.1540) 0.1564 (0.1449) 0.2604 (0.0702) 0.7391 0.068 (0.7931)

3.6444 (1.4871) 0.1535 (0.0610) 0.0424 (0.0538) 0.0618 (0.6582) 0.0446 (0.0183) 0.0556 (0.0374) 0.0050 (0.0477) 0.2124 (0.1544) 0.1614 (0.1453) 0.2525 (0.0694) 0.7389 0.072 (0.7879)

Constant Airport HHI

Total passenger volume Mean distance Real wage Population Unemployment rate Adjusted R-squared Sargan–Hansen (p-value)

0.0813 (0.0645) 0.1084 (0.6450) 0.0350 (0.0193) 0.0665 (0.0347) 0.0224 (0.0477) 0.2222 (0.1528) 0.1414 (0.1439) 0.2475 (0.0686) 0.7393 0.043 (0.8349)

Notes: 1. Method used – airport-level fixed effects two-stage GLS. One-year lags are used as instruments for HHI, highest market share, route HHI, and passenger volume. 2. Reported results are for sub-sample of primary airports handling less than 0.05% of all passenger boardings within the US in a given year. 3. All specifications use natural logarithms of all independent variables reported here, except for constant. 4. Standard errors, robust to heteroscedasticity across and autocorrelation within cross-sections, are reported in parentheses. 5. Airport-level airline market shares and year dummy variables are included in all specifications, but not reported. 6. Included observations: 3742. 7. Significance: 10%; 5%.

high, especially for the sub-sample of smaller airports. This leads us to suggest that there is a possibility that, despite the fact that we used a rather long panel, variation of route-level concentration within airports has simply not been sufficient for us to observe any appreciable effect of route-level competition on airport-level airfares in the panel data setting. To put things into clearer perspective, consider the following fact. The average coefficient of variation for within-airport route-level concentration in our data is a bit more than 8%.14 The same measure for airport-level HHI is close to 60%. Thus, there is simply more within-variation in airport-level concentration than in the route concentration to enable potential identification of the corresponding effects. At the same time, the facts above make it even more remarkable that we have been able to observe a strong effect of route-level competition on airfares in the sub-set of larger airports.

6. Application – two recent and one projected merger 6.1. Impact of mergers on airport concentration Our data analysis results may have interesting implications for the analysis of airline consolidation. Three high-profile mergers occurred in the US airline industry over the last five years. In 2008 Delta Air Lines acquired Northwest Airlines, in 2010 Continental Airlines and United Airlines merged, keeping the latter carrier’s name and redesigning their logo to reflect legacy of both merger partners. The former merger was approved largely thanks to the partner airlines’ complementary networks. A recent study also demonstrated that the extent of route-level competition between the legacy carriers did not considerably affect market-level airfares (Brueckner et al., 2013). In 2011, Southwest Airlines bought AirTran – a fellow LCC based in Atlanta. Additionally, at the time of this writing, a merger between American Airlines and US Airways is in the works. This proposed merger has been challenged by the US Department of Justice, with the trial set for November of 2013.15 It is generally true that individual non-stop routes in the US airline industry are quite concentrated – our data clearly demonstrates this. Such high concentration at the route level is related predominantly to the fact that most airlines operate hub-and-spoke networks. Thus, many if not most of the non-stop flights out of smaller airports link these airports with the 14

We computed coefficients of variation for this variable for each individual airport, and then calculated a simple non-weighted average of these coefficients. Interestingly, an earlier version of our paper has been noted in the US House of Representatives hearing on the proposed American-US Airways merger, on February 26, 2013. 15

299

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

respective airlines’ hubs. Then, while both recent and proposed mergers have not affected the extent of non-stop competition on most routes – except for markets connecting the merger partners’ hub airports – competition on one-stop markets has decreased (see the next sub-section for more details on this point). Another likely outcome of the airline mergers is increased concentration at the airport level, especially at smaller airports served by many of the major carriers to feed passenger traffic to the respective airlines’ hubs. Our regression results demonstrate a clear relationship between airport level concentration and average fares for trips originating at small hubs and non-hub airports. In this sub-section, we will try to quantify the consumer welfare implications of the Delta-Northwest, United-Continental, and projected American-US Airways mergers, suggested by our estimation results. To evaluate the projected changes in airport-level concentration, we have taken the corresponding data (airport-level passenger-based HHI and merger partners’ airport-level market shares) for 2007 in the case of Delta-Northwest and 2009 for the United-Continental and American-US Airways mergers. We assumed, quite naively, admittedly, that the post-merger airport-level market share would be equal to the sum of the partners’ pre-merger numbers, everything else being equal. Then, the change in the HHI as a result of the merger would be computed as:

DHHI ¼ 2Si Sj

ð1Þ 16

where Si and Sj are the partner airlines’ pre-merger market shares. We have then proceeded to identify the airports, where a merger would result in a 5 or higher percent increase in passenger-based airport-level HHI.17 For the Delta-Northwest merger, 54 such airports have been identified. The corresponding number for the Continental-United consolidation event was only 14. The American-US Airways merger is expected to increase airport-level HHI by more than 5% in 29 airports. Further, three airports (Boston Logan, Los Angeles, and Orlando) are projected to be hit by at least 5% increase in HHI as a result of each of the three consolidation events. 17 airports are projected to be hit with at least a 5% increase in HHI from two out of three mergers. Our analysis will be based on the premise that the observed relationship between prices and airport-level HHI was found to be driven by those gateways that were classified as small hubs or non-hub primary airports, according to the share of total US domestic traffic handled. This consideration means that both the Continental-United and American-US Airways consolidation events are not projected to increase airport-level fares via the airport-concentration–price link. Specifically, out of the 14 airports which we identified as candidates for increased concentration level following the Continental-United merger, only 3 belonged to this category. Also, projected average price increase following the merger for flights originating at those airports was only between 0.8% and 1.4%. The projected American-US Airways consolidation event will lead to a significant increase in HHI only in 2 small hub airports. All other airports, concentration at which is projected to be adversely affected by this merger, are either large or medium hubs. The effects of the Delta-Northwest merger, however, warrant further investigation, as we determined that 32 small hub and non-hub airports, handling nearly 20 million passengers in total in 2007 (this corresponds to about 3% of the total domestic US passenger traffic in that year), would become significantly more concentrated after this merger. As can be seen from Table 8, 9 of those small airports were projected to experience 50% or higher increase in HHI, which would, according to our estimates, translate in price increases of up to 11% for trips originating at those gateways. To evaluate the projected percentage changes in price levels corresponding to the projected post-merger change in HHI, we used the corresponding regression coefficients from specifications (1) and (6) in Table 5 (recall that this is the table that reports results for the sub-sample of small hub and non-hub airports). Note that these coefficients respectively represent the smallest and the largest estimated effect of airport concentration on prices. The corresponding coefficients from Table 2 are close in magnitude to the lower end of the values we employ for our exercise. Then, assuming a unitary price elasticity of demand for air travel (according to Brons et al., 2002, a unitary price elasticity lies in the middle of the range of values for price elasticity of demand for air travel, as reported by various studies), change in consumer surplus can be approximated through:18

DCS ¼ Q 0 P0 ð%DPÞð1  0:5ð%DPÞÞ

ð2Þ

where Q0 and P0 correspond to pre-merger passenger volume and average price, taken directly from our dataset. The term%DP denotes percentage change in price, computed by multiplying the corresponding elasticity estimate from the regression results by the projected change in airport-level HHI. We should note, however, that our results are not that sensitive to the choice of price elasticity of demand. Indeed, repeating our calculations with a demand elasticity of 0.5 increases the estimated consumer welfare loss by less than 2%. Table 7 reports our estimates of consumer welfare losses following the Delta-Northwest merger for each of the 32 small airports, which were projected to experience more than a 5% increase in the value of the airport-level concentration measure (passenger based HHI). Over these 32 airports, we discover consumer welfare losses totaling about one quarter of a billion dollars per year. In all honesty, this number should be halved, as the effect we observed applies to trips originating at an airport. Then, it is reasonable to assume that half of the passenger traffic reported in Table 7 corresponds to passengers orig16 Without loss of generality, assume airlines 1 and 2 of N present on the market merge. Then, pre-merger HHI can be written as HHI0 ¼ S21 þ S22 þ P whereas post-merger HHI is HHI1 ¼ ðS1 þ S2 Þ2 þ Ni¼3 S2i ; the subsequent algebra is trivial. 17 Five percent increase in airport-level HHI would correspond to about 0.5% increase in average airfare. 18 This calculation assumes demand relationship is linear or near linear, and that demand curve itself does not shift following the merger.

PN

2 i¼3 Si ;

300

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

Table 6 Individual airline market share effects.

American Airlines Alaska Airlines JetBlue Airways Continental Airlines Delta Air Lines Frontier Airlines Air Tran Airways America West Airlines Northwest Airlines TWA United Airlines US Airways Southwest Airlines

All airports

Large and medium hubs

Medium and small hubs

Small hubs and non-hub

0.3115 (0.1038) 0.2058 (0.0618) 0.4837 (0.2384) 0.0700 (0.1032) 0.1346 (0.0813) 0.0673 (0.1345) 0.0587 (0.1500) 0.0182 (0.1447) 0.0570 (0.1130) 0.1636 (0.1096) 0.0132 (0.1553) 0.3204 (0.0911) 1.1920 (0.2554)

0.1294 (0.4784) 0.5119 (0.5409) 0.8894 (0.5685) 0.4452 (0.4227) 0.5790 (0.4518) 0.6616 (0.3014) 0.1026 (0.2908) 0.7013 (0.4593) 1.5333 (0.7122) 1.2998 (0.5176) 0.1576 (0.4192) 1.2959 (0.4809) 2.6746 (0.5034)

0.5943 (0.2287) 0.1113 (0.0632) 1.1425 (0.5461) 0.1646 (0.1585) 0.1080 (0.1969) 0.0193 (0.2356) 0.0042 (0.2624) 0.0104 (0.2453) 0.7281 (0.2572) 0.3180 (0.1913) 0.0766 (0.2785) 1.1440 (0.2485) 1.4432 (0.3659)

0.3771 (0.0983) 0.1803 (0.0743) 0.3511 (0.2283) 0.0068 (0.0923) 0.1481 (0.0859) 0.0829 (0.1383) 0.0128 (0.1461) 0.0814 (0.1282) 0.0301 (0.0994) 0.1268 (0.1029) 0.0188 (0.1578) 0.2615 (0.0926) 0.7486 (0.2143)

Notes: 1. This table reports coefficients for airline market share variables from regression (4) in Tables 2–5, respectively. 2. Prior to 1997 merger of Valujet Airlines with Airtran Airways, Airtran’s market share corresponds to Valujet’s, as it has been computed using the carrier’s IATA code FL. 3. Some carriers were not present in the data for the entire duration of our panel. In particular: a. JetBlue Airways was founded in 2001. b. Frontier Airlines was founded in 1994. c. America West Airlines merged with US Airways in 2005. d. Northwest Airlines merged with Delta Air Lines in 2008. e. TWA merged with American Airlines in 2000.

inating at a given airport, whereas the other half are passengers whose trips originated elsewhere. We can also safely assume that connecting traffic at the airports listed in Table 7 is minimal if at all existent. The figure for consumer welfare loss we have come up with, while in hundreds of millions dollars per year, appears miniscule if put into context of the multi-billion-dollar US airline industry. According to the US DOT, total operating revenue of US airlines on the US domestic market in 2007 amounted to nearly $130 billion dollars. This means that the welfare loss we computed corresponds to about one tenth of 1% of the US airlines’ total annual revenue. Our results, however, suggest that the adverse effects of the merger might be concentrated in a handful of metropolitan areas, rather than being dispersed across the entire US domestic airline industry. In fact, an average traveler starting his/her journey from either of the 32 small airports identified to experience price increases as a result of the merger, is expected to experience nearly $13 of welfare loss following the consolidation event, after an expected $13.29 passenger weighted average increase in airfare for travel originating at those airports. Price changes and per traveler welfare losses of this magnitude, occurring across the whole airline industry, would probably make front page headlines. Of course, our estimate of consumer welfare effects of the Delta-Northwest merger across a number of small airports is naïve and does not take account of possible changes in the airlines’ conduct post-merger, as well as of generally increased degree of price competition in the industry over recent years – see Berry and Jia (2010). Peters (2006) also shows that naïve approaches to merger simulations did not yield satisfactory results when applied to a series of 1980s airline consolidation events. Indeed, following a merger, we could see an overall reduction of service frequency by the merged airline, along with entry or expansion by other carriers, which would suggest that we might have over-estimated the projected increases in airport-level HHIs. Further, we have effectively assumed that any potential efficiency gains due to the merger would not be passed along to the passengers. Our evaluation presented here could nevertheless serve as a baseline, on which future studies may elaborate. 6.2. Impact on route-level concentration As for the other two consolidation events we considered (the United-Continental merger and the projected American-US Airways merger), we have previously claimed that those events are unlikely to yield significant increases in average

Table 7 Welfare effects of Delta-Northwest merger in small hubs and non-hub airports. Percentage change in airport HHI

Percentage change in price, low limit

Price change, high limit

Passenger enplanements, 2007

Average airfares, year 2007 prices

AVP AZO BIS CID DLH FAR FCA FNT FSD GFK GTF HDN LAN LSE MBS RAP RST TVC ALB BHM BTR DSM GPT GRB GRR MDT MSN PWM ROC SRQ SYR VPS

41.16 13.17 55.61 6.86 79.46 44.21 6.36 32.13 48.20 63.43 10.25 8.88 76.17 63.11 68.23 12.61 33.20 64.08 8.15 5.51 26.28 8.42 22.76 68.91 69.24 10.31 38.23 9.78 10.53 9.24 22.74 66.09

4.61 1.47 6.23 0.76 8.90 4.95 0.71 3.60 5.40 7.11 1.14 0.99 8.53 7.07 7.64 1.41 3.72 7.18 0.91 0.61 2.94 0.94 2.55 7.72 7.76 1.15 4.28 1.09 1.18 1.03 2.54 7.40

5.79 1.85 7.83 0.96 11.18 6.22 0.89 4.52 6.78 8.93 1.44 1.25 10.72 8.88 9.60 1.77 4.67 9.02 1.14 0.77 3.70 1.18 3.20 9.70 9.74 1.45 5.38 1.37 1.48 1.30 3.20 9.30

220,190 189,276 183,787 531,864 175,429 303,563 196,839 534,774 399,533 85,196 190,203 139,573 257,893 123,753 187,290 239,642 166,853 202,330 1,423,672 1,810,455 485,765 969,290 452,485 447,344 1,001,098 644,469 789,832 819,997 1,428,543 775,056 1,184,155 381,477

421.14 471.01 455.54 406.69 416.42 492.84 532.74 281.36 433.12 564.55 489.11 475.21 364.39 465.38 441.07 476.61 378.20 479.08 337.41 336.62 412.80 394.75 362.06 401.46 420.67 433.27 424.99 367.71 304.50 288.80 359.16 547.21

Total consumer welfare loss

Consumer welfare loss, low limit

Consumer welfare loss, high limit

4,670,498 1,462,524 5,644,080 1,857,934 6,927,760 8,077,667 835,350 5,950,285 10,161,107 3,679,351 1,190,118 735,766 8,561,029 4,384,816 6,776,089 1,795,177 2,576,998 7,487,977 4,895,670 4,209,763 6,511,866 4,025,400 4,618,015 14,871,345 35,033,699 3,593,129 15,732,587 3,684,361 5,716,197 2,584,025 11,976,917 16,610,941

5,674,326 1,784,089 6,842,368 2,268,460 8,367,859 9,809,360 1,019,999 7,238,775 12,332,109 4,455,218 1,452,390 898,084 10,345,955 5,309,699 8,198,936 2,190,055 3,134,536 9,066,071 5,976,321 5,140,919 7,928,695 4,913,768 5,625,647 17,992,157 42,383,484 4,384,927 19,122,253 4,496,596 6,975,625 3,153,928 14,590,276 20,105,496

216,838,457

263,178,397

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

Airport code

Note: we only include airports, for which projected increase in HHI following the merger was higher than 5% from the pre-merger level. Estimates of consumer welfare loss are based on passenger HHI coefficient from Table 5, specifications (1) and (6) for low and high limit, respectively. Unitary demand elasticity is assumed in all calculations. Locations of airports listed in Appendix A.

301

302

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

airport-level prices via changes in the airport-level concentration. This finding is based on the assertion that (a) the airportconcentration–price relationship is driven by small airports, (b) the majority of the airports projected to become more concentrated following these mergers are medium and large hub airports, and that (c) given the generally non-overlapping structure of the merger partners’ networks, we cannot expect substantial changes in the route-level competition following these mergers, at least at the non-stop flights level. At the same time, high route-level concentration for the direct flights might still translate into sizeable competition on the markets where only one-stop services are feasible, given that the airlines on the US market operate hub-and-spoke networks. To shed light on this issue, we have been able to obtain information on airlines’ market shares for 2008 from the US DOT website, computed from the original DB1B data. DB1B dataset is a sample of actual itineraries used by passengers on the US domestic market. This means that the information complied by the US DOT from that dataset will, inter alia, allow us to evaluate the extent of market-level competition between the airlines on one-stop routes. The DOT provided this summary data for markets involving a medium or large hub airport with at least 10 passengers per day. In the end, we were able to compute passenger-weighted route-level HHI by the origin airport for 232 gateways, including of course all the airports classified as large and medium hubs. This fact is very convenient for us, as we have concluded that route-level concentration affects average airfares for trips commencing at an airport only in the sub-sample of large and medium hubs. The mean value of this route-level HHI across all the airports is 0.59 with a standard deviation of 0.20 – implying significantly higher market-level competition than what is inferred by the route-level HHI based on non-stop flights.19 Based on the passenger-weighted airline market shares as computed by the US DOT, we have evaluated the changes in the route-level HHI following each of the three mergers. We have once again taken a naïve approach of adding the merger partners’ market shares on each individual route, and recalculating the passenger-weighted route-level HHI based on this assumption. The route-level HHI, averaged across the 232 airports, increased very little: the Continental-United merger is projected to yield a 0.2% increase in this measure, and the corresponding figures for Delta-Northwest and American-US Airways airports are 0.6% and 1.6%, respectively. The numbers for the projected hike in route-level HHI are somewhat higher for the sample of large and medium hub airports: 0.6%, 1.0%, and 2.1% increase in HHI for the Continental-United, Delta-Northwest, and AmericanUS Airways mergers, respectively. We observe substantial differences in the projected effect of the mergers on route-level HHI, both across the airports and the mergers. Focusing on large and medium hub airports, we detect 11 gateways, for which the American-US Airways merger will be projected to increase route-level HHI by 5% or more.20 The corresponding number for the Continental-United merger is only 2 – Newark Liberty and New Orleans. The Delta-Northwest merger is projected to increase route-level HHI by over 5% in only 1 hub airport – Memphis. This is quite in contrast to the non-trivial number of small hub and non-hub airports where this merger was projected to substantially increase airport-level concentration. We then computed the projected change in consumer surplus following each of the three events, using 0.7 as the price elasticity of route-level concentration (this is approximately the average of the four relevant coefficients from Table 3). We have used the same methodology as outlined in the previous section, with one small modification: passenger volumes were estimated from the aggregated DB1B data, rather than taken from the T-100 dataset. We have done this to properly account for true origin-and-destination as opposed to connecting passengers. This is an especially important issue for the hub airports. For instance, only about 23 million of over 80 million passengers handled at Atlanta airport (ATL) constituted traffic originating or terminating at ATL – the rest were just using this airport to change planes. Table 8 presents results of our calculations for the 14 airports, where any of the three mergers is expected to increase passenger-weighted mean route-level HHI by at least 5%. We see that the proposed American – US Airways merger is responsible for substantial potential welfare losses through higher route-level HHI. Across all large and medium hub airports, our calculations yielded nearly $3 billion worth of consumer surplus lost by the passengers as a result of the three consolidation events. The corresponding results are not provided to save space, but available from the authors upon request. Notably, about 60% of this loss was attributed to the projected American-US Airways merger, with Continental-United and Delta-Northwest events each responsible for about 20% of the consumer welfare loss. The following facts should also be noted. First, the reported consumer welfare losses represent only $5.20 per passenger projected loss as a result of the American-US Airways merger; the numbers for the other two mergers are about $1.80–1.90 per passenger. At the same time, per passenger losses at hub airports, where consolidation is projected to increase route-level HHI by more than 5% constitute about $20 per origin–destination passenger (with Charlotte, NC topping the list with $31 lost per passenger as a result of the projected American-US Airways consolidation event). In the end, projected welfare losses from the mergers via increase in route concentration appear to be concentrated at a handful of large and medium hub airports.21 Interestingly, many of the airports adversely affected by the consolidation events here are the merger partners’ hubs (not all such hub airports however are hit by an increase in route-level HHI22). As before,

19 The corresponding statistics are only available for several years in our panel, which does not allow robust estimation of our specifications using this routelevel HHI measure. 20 These airports are: Hartford, CT; Columbus, OH; Raleigh-Durham, NC; Richmond, VA; San Antonio, TX; Charlotte, NC; Washington Regan National (DCA); Dallas-Ft. Worth, TX; Miami, FL; Chicago-O’Hare, IL; and Philadelphia, PA. 21 For instance, 11 hub airports, where route-level HHI is projected to increase by more than 5% following the American-US Airways merger are responsible for nearly three quarters of all projected consumer welfare loss following this event. 22 Examples from the mergers considered here include Detroit, Minneapolis, Salt Lake City, Denver, San Francisco, and Houston Intercontinental.

303

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305 Table 8 Large and medium hub airports with substantial changes in route-level HHI. Airport

Hartford, CT Columbus, OH Memphis, TN New Orleans, LA Raleigh-Durham, NC Richmond, VA San Antonio, TX Charlotte, NC Washington, CD (Regan National) Dallas-Ft.Worth, TX Newark, NJ Miami, FL Chicago, IL (O’Hare) Philadelphia, PA Total

Percentage change in route HHI

Consumer welfare loss

AA-US

CO-UA

DL-NW

AA-US

CO-UA

DL-NW

5.14 5.62 2.91 0.13 10.1 8.10 5.71 11.9 5.09 9.28 0.27 8.35 10.6 9.92

0.00 0.39 0.11 7.02 0.00 3.40 2.03 0.00 0.11 0.01 7.71 0.23 0.00 0.00

1.14 1.93 6.99 0.73 1.32 0.00 3.23 0.70 0.80 0.15 0.24 0.34 0.22 0.37

59,844,304 50,237,115 8,460,340 425,889 22,736,547 5,895,566 9,384,282 175,485,079 123,722,480 350,774,479 8,932,300 51,303,058 203,289,209 101,588,099

0 3,584,796 310,770 22,943,105 0 2,516,073 3,372,766 0 2,693,171 516,951 251,326,293 1,446,378 0 0

13,515,778 17,457,711 20,046,702 2,447,093 3,056,938 0 5,357,345 10,797,824 19,684,523 5,899,508 7,935,217 2,162,274 4,300,666 3,934,986

1,172,078,747

288,710,305

116,596,565

Note: This table includes consumer welfare loss calculations for fourteen airports, where any of the three airline mergers are projected to yield at least 5% increase in mean route-level HHI. Route-level HHI was calculated based on aggregated DB1B data for 2008, taking into account both non-stop and one-stop routes originating at the airports. All calculations assume unitary price elasticity of demand and 0.7 price elasticity with respect to route-level HHI, based on estimates from Table 3. Mean airfares and passenger volumes are based on DB1B data.

however, our assessment of the effects of the merger is admittedly naïve, as we do not take into account the possibility of the partner airlines consolidating their services or shutting down some of their hub airports (see Bilotkach et al., 2013, for an analysis of the network reorganization following the Delta-Northwest merger). Additionally, we apply the estimate based on routelevel concentration measures at the non-stop market level to the data for both non-stop and one-stop flights. It is difficult to say at this time whether the magnitude of the price-route-concentration relationship would be the same if the data we used for our evaluation exercise were available to us for the entire panel. We are leaving this issue for future research. Despite all the potential problems we have outlined above, we can say that our research contributes to understanding effects of consolidation in the airline industry by pointing out the possibility of unequal distribution of gains and losses due to mergers. Specifically, our findings imply the possibility of non-trivial consumer welfare losses in smaller communities, and may be related to increased concentration at the respective airports, even where competition on individual nonstop routes might not be affected by the consolidation event. An interesting question for further study is whether increased airport-level concentration leads to higher fares for trips originating at an airport via decreased competition on one-stop markets.

7. Conclusions Our study differs from most papers examining pricing in the airline industry in one important way. Instead of looking at prices at the airline-route level, we examine determinants of an average airfare for trips originating at an airport. From the literature on pricing in the airline industry, we can identify at least four possible drivers of the airport-concentration and airport-level airfare relationship, which are largely independent of the non-stop route-level competition between airlines. First, airlines operating hub-and-spoke networks will inevitably compete on one-stop routes that originate at the airport, flying passengers to the same destinations via different hubs. Second, airlines enjoying a dominant position at an airport can influence an airport’s decision making, including blocking further entry. Third, an airline’s entry into an airport can be construed by incumbent carriers as a competitive threat, affecting fares on routes not directly served by the entrant. Finally, we can expect airlines at less concentrated airports to compete more aggressively for frequent fliers residing in the airport’s catchment area. Examining price setting at the airport level allows us to explore whether and to what degree airport-level concentration, competition on routes originating at an airport, and the largest carrier’s market share will affect the airfare an average passenger will pay. We also take a comprehensive look at the US airline industry, examining price setting at both large and smaller airports – in this way, we differ from many studies which choose to focus on larger airports and/or denser markets. Overall, our results demonstrate the following. First, effects of airport-level concentration on airfares for trips originating at an airport depend on the airport size. Specifically, the relationship is the strongest in smaller airports. There is also some evidence suggesting that airport concentration might affect fares in the largest airports. In mid-sized airports, however, concentration does not appear to affect average fares for trips originating at those gateways. Second, route-level concentration appears to be a robust predictor of airport-level fares at larger airports. Third, the pure airport dominance effect (as proxied by the airport-level market share of the largest carrier) does not appear to be robust enough to including airport-level and/or route-level concentration measures into specifications.

304

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305

Our study leaves a number of important questions open. First, while we list potential reasons behind the airport-concentration–price relationship, we do not approach the question of which of these reasons is/are responsible for the result we obtained. Addressing this issue would require access to more detailed data on competition on one-stop markets, airport-airline relationships, and frequent flier program membership, among other things. Second, we only look at average fares for trips originating at an airport. Whereas previous studies have demonstrated that, for instance, magnitude of the airport dominance premium can differ at the upper end of the price distribution, as price insensitive travelers typically tend to be more frequent fliers, who value their preferred airline’s frequent flier program more. Third, some of the results, as we indicated above, can be related to and potentially even driven by the extent of within-variation in the key concentration measures. For example, route competition may have an effect on prices at both larger and smaller airports; however, the limited within-variation in the extent of such competition at smaller gateways might have led us to conclude that such an effect is not statistically significant. The take-away message from our study is that sources of market power appear to be different for trips originating at airports of different size. Also, the consolidation currently being observed in the US airline industry can have different effects on travelers living near larger and smaller airports, which means estimates of industry-wide effects of changes in the competitive environment in the US airline industry should be taken with an understanding that gains and losses might not be distributed equally. In particular, our analysis points to a number of small communities that may end up with non-trivial consumer welfare losses following recent mergers. Acknowledgements We thank Philip Gayle, Jan Brueckner, conference participants in Sacramento, Arlington, and Washington, and seminar participants at Leeds University for helpful feedback. Comments by an anonymous Referee have been instrumental in improving this paper. Appendix A A.1. List of airports affected by the Delta-Northwest merger

Airport code

Location

AVP AZO BIS CID DLH FAR FCA FNT FSD GFK GTF HDN LAN LSE MBS RAP RST TVC ALB BHM BTR DSM GPT GRB GRR MDT MSN

Scranton, Pennsylvania Kalamazoo, Michigan Bismarck, North Dakota Cedar Rapids, Iowa Duluth, Minnesota Fargo, North Dakota Glacier Park, Montana Flint, Michigan Sioux Falls, South Dakota Grand Forks, North Dakota Great Falls, Montana Hayden, Colorado Lansing, Michigan La Crosse, Wisconsin Saginaw, Michigan Rapid City, South Dakota Rochester, Minnesota Traverse City, Michigan Albany, New York Birmingham, Alabama Baton Rouge, Louisiana Des Moines, Iowa Gulfport-Biloxi, Mississippi Green Bay, Wisconsin Grand Rapids, Michigan Harrisburg, Pennsylvania Madison, Wisconsin

305

V. Bilotkach, P.A. Lakew / Transportation Research Part A 59 (2014) 288–305 List of airports affected by the Delta-Northwest merger (continued)

Airport code

Location

PWM ROC SRQ SYR VPS

Portland, Maine Rochester, New York Sarasota, Florida Syracuse, New York Fort Walton Beach, Florida

References Bailey, E.E., Panzar, J.C., 1981. The contestability of airline markets during the transition to deregulation. Law Contemp. Probl. 44 (1), 125–145. Bailey, E.E., Graham, D.R., Kaplan, D.P., 1985. Deregulating the Airlines. The MIT Press, Cambridge, MA. Bel, G., Fageda, X., 2008. Getting there fast: globalization, intercontinental flights, and location of headquarters. J. Econ. Geogr. 8, 471–495. Berry, S., 1990. Airport presence as product differentiation. Am. Econ. Rev. Pap. Proc. 80, 394–399. Berry, S., Jia, P., 2010. Tracing the woes: an empirical analysis of the airline industry. Am. Econ. J.: Microecon. 2, 1–43. Bilotkach, V., 2007. Asymmetric regulation and airport dominance in international aviation: evidence from the London-New York market. South. Econ. J. 74, 505–523. Bilotkach, V., 2011. Multimarket contact and intensity of competition: evidence from an airline merger. Rev. Ind. Organ. 38, 95–115. Bilotkach, V., Hueschelrath, K., Mueller, K., 2012. The construction of a low-cost airline network: facing competition and exploring new markets. Manag. Decis. Econ. 33, 485–499. Bilotkach, V., Fageda, X., Flores-Fillol, R., 2013. Airline Consolidation and the Distribution of Traffic between Primary and Secondary Hubs, Working Paper. Boguslaski, C., Ito, H., Lee, D., 2004. Entry patterns in the southwest airlines route system. Rev. Ind. Organ. 25, 315–340. Borenstein, S., 1989. Hubs and high fares: dominance and market power in the U.S. airline industry. Rand J. Econ. 20 (3), 344–365. Borenstein, S., 1990. Airline mergers, airport dominance, and market power. Am. Econ. Rev. Pap. Proc. 80, 400–404. Borenstein, S., 1991. The dominant-firm advantage in multiproduct industries: evidence from the U.S. airlines. Quart. J. Econ. 106 (4), 1237–1266. Borenstein, S., 2011. What Happened to Airline Market Power?, Working Paper. Brons, M., Pels, E., Nijkamp, P., Rietveld, P., 2002. Price elasticities of demand for passenger air travel: a meta-analysis. J. Air Transp. Manage. 8, 165–175. Brueckner, J.K., Lee, D., Singer, E., 2013. Airline competition and domestic U.S. airfares: a comprehensive reappraisal. Econ. Transport. 2, 1–17. Ciliberto, F., Williams, J.W., 2010. Limited access to airport facilities and market power in the airline industry. J. Law Econ. 53, 467–495. Clougherty, J., 2002. US domestic airline mergers: the neglected international determinants. Int. J. Ind. Organ. 20, 557–576. Dresner, M., Windle, R., 1992. Airport dominance and yields in the U.S. airline industry. Logist. Transport. Rev. 28, 319–320. Evans, W.N., Kessides, I.N., 1993. Localized market power in the U.S. airline industry. Rev. Econ. Stat. 75, 66–75. Fruhan, W.E., 1972. The Fight for Competitive Advantage: A Study of the United States Domestic Trunk Carriers. Harvard University Graduate School of Business Administration, Boston, MA. Goolsbee, A., Syverson, C., 2008. How do incumbents respond to the threat of entry? Evidence from the major airlines. Q. J. Econ. 123, 1611–1633. Hurdle, G.J., Johnson, R.L., Joskow, A.S., Werden, G.J., Williams, M.A., Johnson, L., 1989. Concentration, potential entry, and performance in the airline industry. J. Ind. Econ. 38 (2), 119–139. Kim, E.H., Singal, V., 1993. Mergers and market power: evidence from the airline industry. Am. Econ. Rev. 83, 549–569. Kwoka, J., Shumilkina, E., 2010. The price effect of eliminating potential competition: evidence from an airline merger. J. Ind. Econ. 58, 767–793. Lederman, M., 2008. Are frequent-flyer programs a cause of the ‘‘Hub Premium’’? J. Econ. Manage. Strategy 17, 35–66. Lee, D., Luengo-Prado, M., 2005. The impact of passenger mix on reported ‘‘Hub Premiums’’ in the U.S. airline industry. South. Econ. J. 72 (2), 372–394. Luo, D., 2013. The price effects of the delta/northwest airline merger. Review of Industrial Organization, in press, http://dx.doi.org/10.1007/s11151-0139380-1. Morrison, S.A., 2001. Actual, adjacent and potential competition: estimating the full effect of southwest airlines. J. Transp. Econ. Policy 35, 239–256. Morrison, S.A., Winston, C., 1995. The Evolution of the Airline Industry. Brookings Institution Press. Peters, C., 2006. Evaluating the performance of merger stimulation: evidence from the U.S. airline industry. J. Law Econ. 49, 627–649. Richard, O., 2003. Flight frequency and mergers in airline markets. Int. J. Ind. Organ. 21, 907–922. Tretheway, M.W., Kincaid, I.S., 2005. The effect of market structure on airline prices: a review of empirical results. J. Air Law Commer. 70 (3), 466–498. U.S. General Accounting Office, 1990. Airline Competition: Higher Fares and Reduced Competition at Concentrated Airports. Report. Van Dender, K., 2007. Determinants of fares and operating revenues at US airports. J. Urban Econ. 62, 317–336.