Abstract
This paper revisits the topic of the effect of airline competition on service quality, but allowing for nonlinear effects. Using a panel of monthly data for 5472 route-carrier combinations from 2005:4Q through 2012:4Q, we find that the average length of flight delays and cancellation rates increase with the concentration level. Worse service quality is linked to less competition. In addition, we find that the relationships between our measures of service quality and market concentration are nonlinear, so that the scale of the effects of a given change in airline competition appears to depend on the initial level of competition.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Earlier studies analyzing the relationship between market structure and product quality also reached various conclusions (Spence 1975, 1976). White (1976) finds an inverse relationship between the number of banking offices and market concentration. White (1977) and Mussa and Rosen (1978) show that price discrimination can affect product quality.
We choose arrival delay with respect to schedule as our measure of flight delay because our focus is on the quality of service as experienced by airline passengers. Consumers only observe scheduled travel time and delays with respect to schedule; they cannot readily get a handle on delays with respect to the shortest travel time. Further, consumer concerns about delays are as likely to focus on reliably reaching a destination by a specific time as they are about having a quick trip. In the robustness section, we discuss alternative flight delay measures.
Congressional interest led the General Accountability Office to report on airline competition issues in 2014, and the Federal Aviation Administration Modernization and Reform Act of 2012 required the U.S. Department of Transportation Office of Inspector General to assess the effects of limited airline service options on the frequency of delays and cancellations.
Well known papers include Borenstein (1989, 1990) on dominance’s giving airlines power to charge higher fares. Borenstein and Rose (1994) and Gerardi and Shapiro (2009) explore the link between airline competition and price dispersion. Yet another group of airline papers focuses on the price effects of low-cost carrier entries; see, e.g., Windle and Dresner (1999), Morrison (2001), Goolsbee and Syverson (2008), Brueckner et al. (2012), and Kwoka et al. (2016). The general conclusion reached in these papers is that greater competition is associated with lower ticket prices.
The ASQP data are linked to the operating carrier, while each observation in the DB1B is tied to a combination of an operating carrier as well as a marketing carrier. It is possible that in the same month and on the same route, one operating carrier is working for multiple marketing carriers and vice versa; e.g., code-sharing. About 50 percent of the observations have an operating carrier that is different from the marketing carrier. We merged these two data sets by operating carrier, and assigned a unique marketing carrier based on the following rules: (1) for records with multiple operating carriers that are all linked to the same marketing carrier, we assigned that marketing carrier; and (2) for records with multiple marketing carriers that are linked to the same operating carrier, we assigned the dominant marketing carrier: the one with greater than or equal to 50 percent of the origin and destination passenger share.
The “major” carriers are: ATA (TZ), AirTran (FL), Alaska (AS), American (AA), American Eagle (MQ), Comair (OH), Continental (CO), Delta (DL), Frontier (F9), JetBlue (B6), Northwest (NW), SkyWest (OO), Southwest (WN), US Airways (US), and United (UA).
The “national” carriers are: Atlantic Southeast (EV), ExpressJet (XE), Mesa (YV), Pinnacle (9E), and Virgin America (VX).
We obtained airport capacity utilization data from FAA’s Aviation Performance Metrics Airport Efficiency System, which is not publically accessible. It covers 77 airports, but only 70 within the lower 48 states.
Because the DB1B does not support distinguishing between a non-stop flight and a flight with one stop but no change of plane, our measures of direct flights include both these types of flights.
Only five percent of the passengers in our sample take trips that involve two or more stops, and we do not include these in our analyses.
We also repeated our analyses using an HHI based on airlines’ passenger market share and found similar results.
To ensure sufficient time series variation for estimating each panel, we omitted route-carrier groups with data that span less than 1 year. For data reliability reasons we excluded any route that reported less than eight flights per month, and flights that arrived more than 1 h early or over 6 h late. We also excluded observations with either an origin or destination outside the continental U.S. Finally, we discarded the last 1 percent of data in each tail of the distribution for the key variables in our sample to eliminate outliers. The results produced that included or excluded these data were nearly identical.
All the “percent” variables in our data are expressed in decimal form.
The congestion variables also capture slot-control effects. Congestion or capacity utilization averaged 62 percent for the slot-controlled airports in our sample, and 49 percent for all other airports.
Group 2 contains regional jets with 70–100 seats. Group 3 is made up of narrow-body planes that have more than 100 seats. Group 4 includes all wide-body aircraft.
We adopt BTS’s carrier assignments following each merger in processing our data.
We collected information on airline labor actions from: the Wall Street Journal, the Chicago Tribune, the Bureau of Labor Statistics, CNN, USA Today, NBC news, TribLive, and Highbeam. Four events were identified during our sample, affecting: (1) Northwest Airlines from October to November 2005; (2) Northwest Airlines in November 2006; (3) US Airways in April 2006; and (4) American Airlines from September to October 2012.
This measure is defined as \({{\sqrt {enp_{i1} *enp_{i2} } } \mathord{\left/ {\vphantom {{\sqrt {enp_{i1} *enp_{i2} } } {\sum\nolimits_{k} {\sqrt {enp_{k1} *enp_{k2} } } }}} \right. \kern-0pt} {\sum\nolimits_{k} {\sqrt {enp_{k1} *enp_{k2} } } }}\), where k indexes all airlines, and enp1 and enp2 are the total monthly passengers (“enplanements”) at the two endpoint airports.
The cumulative normal density function of Eq. (2), which has both convex and concave regions, accommodates nonlinearities beyond those that are captured by inclusion of the quadratic form.
To conduct a cluster-robust Hausman test that compares fixed effects (FE) and random effects estimators in the delay length model, we also rely on CRE, following Wooldridge (2010a). The test indicated that we needed to include FE.
We also estimate the linear specifications and find that our results are consistent with previous work (Mazzeo 2003; Rupp and Holmes 2006; Greenfield, 2014) that suggests that competition is associated with reduced delays. In the linear specification estimation, the coefficient of HHIjt is 13.04 and is significant at the 1 percent level. We do not present estimation results for other variables in the linear specification because they are almost identical to those in the nonlinear specification, which are summarized in Table 4 column 1. Our linear specification results also indicate that cancellation rates fall as competition increases.
Note that the CF indicator of endogeneity of the competition measure, v2hat, is significant.
The Tobit model assumes that the functional form is unaffected by whether or not an airline cancels a flight—which may not be appropriate here. Estimation of the marginal effects discussed above included observations with zero cancellations. If instead we condition on a positive cancellation rate, the marginal effects are smaller, ranging from 0.11 % at HHI = 0.1–0.23 % at HHI = 0.9 for a 10 % increase in HHI.
We also use the fitted values of an OLS regression with the same variables as alternative measures of ideal elapsed time. We find that competition has no effect on delays measured in this way.
The average elapsed time delay in our sample was 15.1 min with a standard deviation of 14.1 min.
Other estimation results are available upon request.
Results that use other subsample cutoffs also indicate that the findings are robust, and are available upon request.
We find that using the log of elapsed time as the dependent variable produces consistent results. Furthermore, taking logs is a better way to deal with the potential heteroscedasticity problems. However, in order to compare the carrier fixed effects among different models, we focus on the level of elapsed time instead of logs. The log regression results are available upon request.
See footnotes 6 and 7 for lists of the airlines in each group, and the accompanying text for definitions.
The low cost airlines in our sample are AirTran, ATA, Frontier, JetBlue, Southwest, and Virgin America.
We exclude the “Depart Early” group in the Chow test due to its having only 56 observations.
The subgroup results of the cancellation models are available upon request.
For the route-carrier combinations that report throughout our sample, this indicator variable is always zero. For the combinations that initially appear and then disappear from the sample, it equals one in the period just before attrition.
In April 2010, the U.S. Department of Transportation implemented a rule with respect to the tarmac delays of domestic flights. The rule addresses holding passengers on aircraft on the ground without the opportunity to deplane. It requires that airlines have contingency plans for handling passenger needs during lengthy tarmac delays, and that a tarmac delay not exceed 3 h unless the pilot-in-command determines that there is a safety or security-related impediment to deplaning passengers, or that air traffic control has advised the pilot-in-command that deplaning would significantly disrupt airport operations Violations can result in fines of up to $27,500 per passenger.
The software (Stata) that we used for estimation does not support cluster robust errors in a SUR model. Thus, we report the default standard errors.
References
Bar-Isaac, H. (2005). Imperfect competition and reputational commitment. Economic Letters, 89(2), 167–173.
Borenstein, S. (1989). Hubs and high fares: Dominance and market power in the U.S. airline industry. RAND Journal of Economics, 20(3), 344–365.
Borenstein, S. (1990). Airline mergers, airport dominance, and market power. American Economic Review Papers and Proceedings, 80(2), 400–404.
Borenstein, S., & Rose, N. L. (1994). Competition and price dispersion in the U.S. airline industry. Journal of Political Economy, 102(4), 653–683.
Brueckner, J. K., Lee, D., & Singer, E. S. (2012). Airline competition and domestic US airfares: A comprehensive reappraisal. Economics of Transportation, 2(1), 1–17.
Brueckner, J. K., Lee, D., & Singer, E. S. (2014). City-pairs versus airport-pairs: A market-definition methodology for the airline industry. Review of Industrial Organization, 44(1), 1–25.
Brueckner, J. K., & Luo, D. (2014). Measuring strategic firm interaction in product-quality choices: The case of airline flight frequency. Economics of Transportation, 3(1), 102–115.
Dai, M., Liu, Q., & Serfes, K. (2014). Is the effect of competition on price dispersion non-monotonic? Evidence from the U.S. airline industry. Review of Economics and Statistics, 96(1), 161–170.
Dana, J. D., & Fong, Y. (2011). Product quality, reputation, and market structure. International Economic Review, 52(4), 1059–1076.
Evans, W. N., Froeb, L. M., & Werden, G. J. (1993). Endogeneity in the concentration-price relationship: Causes, consequences, and cures. Journal of Industrial Economics, 41(4), 431–438.
Gaynor, M. (2006). What do we know about competition and quality in health care markets. NBER Working Paper no. 12301.
Gerardi, K. S., & Shapiro, A. H. (2009). Does competition reduce price dispersion? New evidence from the airline industry. Journal of Political Economy, 117(1), 1–37.
Goolsbee, A., & Syverson, C. (2008). How do incumbents respond to the threat of entry? Evidence from the major airlines. Quarterly Journal of Economics, 123(4), 1611–1633.
Greenfield, D. (2014). Competition and service quality: New evidence from the airline industry. Economics of Transportation, 3(1), 80–89.
Kumbhakar, S. C., & Lovell, C. A. K. (2003). Stochastic frontier analysis. Cambridge: Cambridge University Press.
Kwoka, J., Hearle, K., & Alepin, P. (2016). From the fringe to the forefront: Low-cost carriers and airline price determination. Review of Industrial Organization, 48(3), 247–268.
Mayer, C., & Sinai, T. (2003). Network effects, congestion externalities, and air traffic delays: Or why not all delays are evil. American Economic Review, 93(4), 1194–1215.
Mazzeo, M. J. (2003). Competition and service quality in the U.S. airline industry. Review of Industrial Organization, 22(4), 275–296.
Morrison, S. A. (2001). Actual, adjacent, and potential competition: Estimating the full effect of Southwest airlines. Journal of Transport Economics and Policy, 35(2), 239–256.
Mussa, M., & Rosen, S. (1978). Monopoly and product quality. Journal of Economic Theory, 18(2), 301–317.
Papke, L. E., & Wooldridge, J. M. (2008). Panel data methods for fractional response variables with an application to test pass rates. Journal of Econometrics, 145(1–2), 121–133.
Prince, J. T., & Simon, D. H. (2009). Multimarket contact and service quality: Evidence from on-time performance in the US airline industry. Academy of Management Journal, 52(2), 336–354.
Richard, O. (2003). Flight frequency and mergers in airline markets. International Journal of Industrial Organization, 21(6), 907–922.
Rupp, N. G., & Holmes, G. M. (2006). An investigation into the determinants of flight cancellations. Economica, 73(292), 749–783.
Rupp, N. G., Owens, D., & Plumly, L. W. (2006). Does competition influence airline on-time performance? In D. M. Lee (Ed.), Advances in airline economics, Vol. 1: Competition policy and antitrust (1st ed., pp. 251–272). Amsterdam: Elsevier.
Spence, A. M. (1975). Monopoly, quality, and regulation. The Bell Journal of Economics, 6(2), 417–429.
Spence, A. M. (1976). Product selection, fixed costs, and monopolistic competition. The Review of Economic Studies, 43(2), 217–235.
White, L. J. (1976). Price regulation and quality rivalry in a profit maximizing model: The case of bank branching. Journal of Money, Credit and Banking, 8(1), 97–106.
White, L. J. (1977). Market structure and product varieties. American Economic Review, 67(2), 179–182.
Windle, R. J., & Dresner, M. E. (1999). Competitive responses to low cost carrier entry. Transportation Research Part E: Logistics and Transportation Review, 35(1), 59–75.
Wooldridge, J. M. (2010a). Econometric analysis of cross section and panel data (2nd ed.). Cambridge, MA: MIT Press.
Wooldridge, J. M. (2010b). Correlated random effects models with unbalanced panels. Retrieved August 1, 2012, from Michigan State University, Department of Economics Web site. http://econ.msu.edu/faculty/wooldridge/docs/cre1_r4.pdf.
Zellner, A. (1962). An efficient method of estimating seemingly unrelated regressions and tests for aggregation bias. Journal of the American Statistical Association, 57(298), 348–368.
Acknowledgments
The authors are grateful for excellent suggestions from the editor and two anonymous referees, as well as for comments from James Dana, Daniel Greenfield, Nicholas Rupp, and Jon Williams. We thank Steven Smith and John Heimlich for their expert opinions on the U.S. airline industry, and Joshua Ingber for helpful research assistance. All remaining errors are our own. The views expressed herein are those of the authors and do not necessarily reflect the positions of the Federal Aviation Administration, Federal Reserve Bank of Chicago, or U.S. Department of Transportation Office of Inspector General.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Cao, K.H., Krier, B., Liu, CM. et al. The Nonlinear Effects of Market Structure on Service Quality: Evidence from the U.S. Airline Industry. Rev Ind Organ 51, 43–73 (2017). https://doi.org/10.1007/s11151-016-9544-x
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11151-016-9544-x