Do enhancements to loyalty programs affect demand? The impact of international frequent flyer partnerships on domestic airline demand.

(44) The Continental-Northwest alliance is the only one to have involved codesharing and is the only one still in effect. In late 2002, United and US Airways initiated a codesharing and FFP partnership and, in 2003, Delta joined the Continental-Northwest alliance.

(45) Unlike international partnerships, which affect domestic demand only by expanding a domestic airline's FFP, partnerships with domestic airlines may affect demand and fare by both expanding the domestic airline's network and increasing airline substitutability (as each airline's FFP points can now be earned and redeemed on the other's flights). Because domestic partnerships have this additional effect, an investigation of their effects is conducted in a separate paper. See Lederman (2008).

(46) However, much of the variation in international partners comes in the second part of the sample, which makes it quite difficult to separate these two things.

(47) Because much of the variation in international partners comes during the second part of the sample, the increase in demand on routes departing from hubs in the latter part of the sample probably reflects a combination of the enhancement effect and the formation of the domestic partnerships.

(48) Only the Continental-Northwest partnership had codesharing in addition to FFP reciprocity. I have estimated the demand equation excluding Continental and Northwest and the results are quite similar. Therefore, my results are not capturing the impact of their broader alliance.

(49) I compared the information in the OFFGs to the program guides, which are produced by the airlines for members of their FFP. The OFFGs appear to take the partnership information almost directly from the program guides. I could not use the program guides themselves for this information because past years' program guides were not always available.

Mara Lederman, University of Toronto; mara.lederman@rotman.utoronto.ca.

This is a revised version of Chapter 1 of my doctoral dissertation. I am grateful to Susan Athey, Nancy Rose, and Scott Stern for their many comments. I also thank Silke Januszewski, Avi Goldfarb, Ken Corts, Ig Horstmann, Tim Simcoe, the editor Ariel Pakes and two anonymous referees for helpful comments. Severin Borenstein generously provided the DOT DBIA data. I thank Michelle O'Neill at Inside Flyer magazine, Mary Kandel at OAG, Claire Fairfax at Reed Business Information, and Nina Rose at Northwest Airlines for help in assembling data. Financial support from the Kellogg School of Management and the Social Sciences and Humanities Research Council of Canada is acknowledged. All errors are my own. TABLE 1 Summary Statistics

Standard

N Mean Deviation Fare and passenger variables

MEAN FARE 109,594 212.10 103.31

20th PERCENTILE FARE 109,594 140.28 75.56

80th PERCENTILE FARE 109,594 279.00 175.71

PASS 109,594 151.39 457.81 Local dominance variables

DEP_HUB 109,594 0.16 0.37

SH_FLTS 109,594 0.17 0.23

SH_FLTS<0.20 (SHCATO) 109,594 0.75 0.44

0.20<=SH_FLTS <0.40 (SHCAT1) 109,594 0.10 0.30

0.40<=SH_FLTS <0.60 (SHCAT2) 109,594 0.05 0.21

0.60<=SH_FLTS (SHCAT3) 109,594 0.11 0.31 FFP variables

FFP_SCALE (1000s) 109,594 23.17 14.32

FFP_SCOPE 109,594 217.18 96.02 Other product characteristics

DIRECT 109,594 0.18 0.38

FREQUENCY 19,474 5.784 4.39 Instruments

CAR_OTH_DIR 109,594 0.19 0.39

CAR_OTH_CON 109,594 1.07 1.20 Observation is an airline-routing-route-quarter. The sample includes coach tickets on AA, CO, DL, NW, UA, and US. Routes are between the top 30 airports based on year 2000 enplanements. Fare variables are half of fares paid for round-trip tickets ($2001). TABLE 2 Impact of FFP Partnerships on Demand (Scale Variable, Hub Measure) Dependent Variable LOG (PASS)

(1) (2) FARE -0.002 -0.03

(0.000) ** (0.001021) ** FARE * DEP_HUB DIRECT 3.335 3.477

(0.014) ** (0.037) ** DIRECT * FREQ 0.172 0.178

(0.002) ** (0.005) ** FFP_SCALE * DEP_HUB -0.002 0.022

(0.001) (+) (0.003) ** Observations 109,594 109,594 Dependent Variable LOG (PASS)

(3) (4) FARE -0.031

(0.001) ** FARE * DEP_HUB 0.013

(0.002) ** DIRECT 3.448 3.328

(0.035) ** (0.014) ** DIRECT * FREQ 0.203 0.172

(0.006) ** (0.002) ** FFP_SCALE * DEP_HUB 0.014 -0.004

(0.003) ** (0.001) * Observations 109,594 109,594 Standard errors in parentheses. (+) significant at 10%; * significant at 5%; ** significant at 1%. All specifications include airline-origin, airline-quarter, and route-quarter fixed effects. Columns (1) through (4) present OLS estimates. Columns (2) and (3) present 2SLS estimates with FARE and FARE * HUB treated as endogenous. The sample includes coach tickets on AA, CO, DL, NW, UA, and US. Routes are between the top 30 airports based on year 2000 enplanements. TABLE 3 Impact of FFP Partnerships on Demand (Scale Variable, Share Categories) Dependent Variable LOG (PASS)

(1) (2) (3) FARE -0.0296 -0.0306 -0.0308

(0.0010) (0.0013) ** (0.0012) ** FARE * DEP_HUB 0.0128

(0.0017) ** FARE * (0.20 < SH_FLTS 0.0023

< = 0.40) (0.0005) ** FARE * (0.40 < SH_FLTS 0.0089

< = 0.60) (0.0010) ** FARE * (0.60 < SH_FLTS) 0.0108

(0.0013) ** DIRECT 3.4771 3.4483 3.4580

(0.0367) (0.0346) ** (0.0353) ** DIRECT * FREQ 0.1777 0.2030 0.1965

(0.0051) ** (0.0059) ** (0.0054) ** FFP_SCALE * DEP_HUB 0.0151

(0.0031) ** FFP_SCALE * 0.0060 0.0058

(0.20 < SH_FLTS < = 0.40) (0.0018) ** (0.0017) ** FFP_SCALE * 0.0076 0.0044

(0.40 < SH_FLTS < = 0.60) (0.0029) ** (0.0028) FFP_SCALE * (0.60 < SH_FLTS) 0.0250 0.0167

(0.0035) ** (0.0032) ** Observations 109,594 109,594 109,594 Standard errors are in parentheses. (+) significant at 10%; * significant at 5%; ** significant at 1%. All specifications include airline-origin, airline-quarter, and route-quarter fixed effects. All specifications present 2SLS estimates with FARE, FARE * DEP_HUB FARE * CAT1, FARE * CAT2, and FARE * CAT3 treated as endogenous. The sample includes coach tickets on AA, CO, DL, NW, UA, and US. Routes are between the top 30 airports based on year 2000 enplanements. TABLE 4 Impact of FFP Partnerships on Demand (Scope Variable, Hub and Share Categories) Dependent Variable LOG(PASS)

(1) (2) FARE 0.0306 0.0296

(0.0013) ** (0.0010) ** FARE * DEP HUB -0.0127

(0.0017) ** FARE * (0.20 < SH FLTS = < 0.40) FARE * (0.40 < SH FLTS < = 0.60) FARE * (0.60 < SH FLTS) DIRECT -3.4482 -3.4769

(0.0346) ** (0.0368) ** DIRECT * FREQ -0.2030 -0.1777

(0.0059) ** (0.0051) ** FFP_SCOPE * DEP HUB -0.0021

(0.0004) ** FFP_SCOPE * (0.20 < SH FLTS < = 0.40) -0.0007

(0.0002) ** FFP_SCOPE * (0.40 < SH FLTS < = 0.60) -0.0012

(0.0004) ** FFP_SCOPE * (0.60 < SH FLTS) -0.0034