- Email: [email protected]
Legacy carriers fight back: Pricing and product differentiation in modern airline marketing
Journal of Air Transport Management 17 (2011) 130e135
Contents lists available at ScienceDirect
Journal of Air Transport Management journal homepage: www.elsevier.com/locate/jairtraman
Legacy carriers fight back: Pricing and product differentiation in modern airline marketing Tim Hazledine Department of Economics, The University of Auckland, New Zealand
a b s t r a c t Keywords: Airline pricing Low-cost airlines Legacy airlines
The adoption of the low-cost carrier business model has applied competitive pressure on established network or “legacy” carriers, by offering fares at prices that legacy carriers find it difficult to match and still cover their fixed costs. This paper reports how two medium-sized national airlines-Air New Zealand and Air Canada-have coped with the low-cost threat by, in effect, turning their fixed costs into profit centres. Features such as full regional networks, long-haul connections, frequent flyer programs, membership in global alliances, lounges and business class cabins can be bundled into products which can be marketed and sold profitably to business and even some leisure travellers, and which cannot be easily replicated by low-cost carriers. Although not panaceas, the innovations of Air New Zealand and Air Canada to the competition they face in their domestic and trans-border markets demonstrate the possibility of an effective legacy carrier response to the low-cost carrier business model. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction This paper examines how two full service national network carriers e Air New Zealand and Air Canada-have responded to two of the major challenges facing airlines in the 21st century: namely, the continued competitive threat from the low-cost carrier (LCC) business model, and the potentialities of the internet as an instrument for marketing and transacting air carrier services. The analytical context is the “empty core” hypothesis (Button, 2004), to the effect that the large fixed costs incurred by “legacy” carriers make it difficult-perhaps impossible-for them to set prices adequate to cover all their costs without those prices being well above marginal variable costs, and therefore vulnerable to predatory marginal cost pricing by rivals, with this unstable situation presumably exacerbated when the rivals are low-cost carriers without large fixed costs of their own to cover because of their flexible, route profitability-based business model. Some legacy carriers have responded to the LCC threat by stripping out their own fixed costs e either by in effect turning themselves into low-cost carriers on short-and medium haul routes (e.g. Aer Lingus), or by setting up LCC subsidiaries and deploying them as a “fighting-brand” on the leisure-oriented routes most susceptible to LCC competition (e.g. Qantas with its subsidiary Jetstar).1
E-mail address: [email protected]. Air Canada (“Tango”) and Air New Zealand (“Freedom Air”) have both experimented with fighting-brand LCC subsidiaries, neither of which are still in operation. 1
0969-6997/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jairtraman.2010.10.008
In contrast, the strategies of Air New Zealand and Air Canada share a quite different response to the LCC business model. Instead of stripping out fixed costs, they turn them to their advantage. The basic idea of this paper is that legacy carriers’ fixed costs can actually deliver market value; that this value can be enhanced by shrewd and innovative pricing and marketing practices, and that such practices cannot easily be replicated by LCCs and so can deliver a source of competitive advantage to the legacy carriers which may compensate for the LCCs’ lower variable costs. These valuable “fixed” elements in Air New Zealand and Air Canada’s product portfolios include their status as national carriers, the density and coverage of their domestic and international route structures, the penetration of their websites, their membership of global alliances, their FFPs, their lounges and other benefits for high-status travellers, and their business class cabins. 2. Background: market penetration of LCCs Although the market penetration of low-cost carriers did increase quite rapidly over the past two decades, there may now be signs emerging of some sort of equilibrium, in which the overall market share of LCCs has stuck at numbers in the 20e40 percent range, depending on market structure. In the US, the LCC share of the domestic market, in terms of number of passengers increased from 4% in 1990 to 20% in 1999, and thence to 33% in 2008, but the share of the six largest legacy carriers was 63.3% in 1997 and 63.1% in 2005, so that it may be that much of increase in LCC penetration is due to the adoption of the LCC business model by most start-ups
T. Hazledine / Journal of Air Transport Management 17 (2011) 130e135
and other small airlines, rather than to a push-back of the positions of the major legacy suppliers.2 Note too that legacy carriers fly longer routes, on average, and so their RPM (revenue passenger miles) share is substantially larger, and their overall revenue share would be larger still, since their fares tend to be higher than those of the LCCs.3 Now, it could be argued that holding onto to market share is not the point e what matters is the cost of so doing. A succession of studies, e.g. Dresner et al. (1996) and most recently Goolsbee and Syverson (2008), have documented the dramatic impact on legacy carrier prices on routes impacted by actual or even just increased probability of entry by LCCs; in particular Southwest. Market share defended by slashing prices when costs are not slashed will impact profitability, and may not be sustainable, as Tretheway (2004) has argued. Nevertheless, the Big-6 (now Big-5) US legacy carriers have chosen to respond to LCC competition with lower prices rather than just by ceding market share, and we should not presume that they don’t have their reasons for doing so. As for profitability e well, the airline industry in North America has somehow survived for decades with long-run rates of return apparently well below what would be consider acceptable in most other sectors. Aviation is a glamorous business, and it may be that the “normal” rate of return on capital is lower in this sector, as it seems to be in other inherently attractive “lifestyle” activities, such as movie production, wineries and even farming in general. The recent spate of Chapter 11 episodes besetting the legacy carriers may in part reflect cash flow problems generated by sharper competition in the marketplace, but probably also has something to do with what could be called the demographic implications of their situation, whereby the incumbents are stuck with a literal legacy of pension and health plan obligations incurred in easier days, as well as a relatively highwage unionised wage structure.4 Berry and Jia (2009) carry out an extremely sophisticated analysis of air travel demand, supply, and profitability in 1999 and 2006, and find that only a quite small proportion e around 10%-of the observed reduction in legacy carriers’ profits between the two years were apparently directly due to the expansion of LCCs over that period, with major factors being greater price sensitivity of customers and a systematic change in preferences towards direct (not connecting) flights. The higher demand price elasticities have been informally linked to greater consumer use of internet ticketing: Dana and Orlov (2009) provide direct evidence that increases in internet penetration can explain the dramatic increase in load factors observed in US markets from just 62%, on average, in 1993 to 80% in 2007. Smaller domestic markets support fewer airlines. In Canada, the near-monopoly created by the folding of the failed CAIL into Air Canada has been challenged successfully by the LCC WestJet, which was launched in 1996 and now takes about 40% of the domestic market as well as some of the trans-border business. In Australia, the failure of Qantas’s long-standing legacy rival Ansett, in 2001, permitted the recently arrived LCC Virgin Blue to expand quickly to
2
These data are taken from tables analysing US Department of Transportation DB1A data, found on www.darinlee.net. The market share of the Big-6 (AA, Continental, Delta, NWA, United, US Airways) was boosted by the absorption of TWA by American in 2002 and of America West by US Airways after 2005. 3 BTS data reported on their RITA website show the six largest legacy carriers holding 47% passenger share and a 58.5% RPM share, in the months up to October 2008, with the same statistics for the three largest LCCs (Southwest, AirTran, JetBlue) being 22.6 and 20.3%,. The discrepancy between these numbers and those reported above may be because the latter uses US Department of Transportation data which do not include all the smallest airlines. 4 A referee has suggested that these legacies may create barriers to down-sizing, which in turn may help explain why the US legacy carriers have fought so hard to retain market share.
131
take around one third of the domestic market e a figure which appears to be acceptable to both airlines and which has been quite stable. In New Zealand, although the inhabitants apparently consume more air miles per capita than in any other country, the size of the market (over four million residents plus over two million annual tourists) has historically been able to support only one profitable carrier (the national incumbent Air New Zealand), despite which first Ansett, then Qantas and now Pacific (Virgin) Blue as well as Qantas have taken advantage of the totally Open Skies regulatory environment to offer domestic services on main trunk and important tourist routes. There are no official data, but Air New Zealand appears to offer more than two thirds of the seats on the routes on which it faces competition, and has a monopoly of the other, regional, routes.5 On the trans-Tasman flights of 3e4 h duration that link the main cities of Australia and New Zealand, the legacy incumbents Air New Zealand and Qantas have since 2002e03 faced competition from Pacific Blue and from Emirates Airlines (making use of 5th Freedom rights), but seem able to hold on to market shares of between 70% and 80% in most markets. 3. The LCC competitive threat: Air New Zealand’s response Although the most obvious competitive challenge posed by the more successful LCCs was the substantially lower price tags on their product, it was the nature of the product itself that may in the longrun have posed the biggest threat to the legacy business model. Indeed, Tretheway judges that: Perhaps the most important impact of the LCC business model on [legacy carriers] has been the introduction of low one-way fares. This has undermined the price discrimination ability of the [legacy carriers], and is the most important pricing development in [the] past 25 years. (2004, p. 5) If the (re)introduction of “low” one-way fares (i.e., not full-fare, unrestricted products aimed at business travellers) is the most important development in a quarter century, this reflects the fact that it challenged what arguably was the most important pricing innovation twenty five years ago e namely, the introduction in the early 1980s by American Airlines of the advance purchase discounted return fare with its key Saturday night stayover (SNS) condition. This brutally simple marketing ploy was of course a clever way of getting high willingness to pay business travellers to self select themselves away from low-WTP leisure customers, with the latter willing and often even keen to secure their ticket well in advance, and undeterred by the SNS condition because most leisure itineraries include a weekend away in any case. The first legacy carrier to introduce a fully worked through response to the LCC business model6 appears to have been Air New Zealand, who in December 2002 rolled out their new ‘Express Fare’ system. The Express system had the following features: The product was pared back, with business class seating taken out, and hot meals and bar service removed, on domestic routes; The previous large number of different fare types or buckets was simplified to three, differing mainly in their refundability
5 Qantas has made no secret of its difficulties in earning profits from its domestic NZ operations. Most recently, the CEO of the airline was reported to have said ‘Qantas came close last year [2008] to exiting the New Zealand domestic market because it was haemorrhaging money’ [See http://www.stuff.co.nz/business/ industries/2924832/Airlines-bleed-in-Tasman-route-scrap]. 6 For a good taxonomy of the differences between legacy and LCC business models, see Gillen and Gados (2008).
132
T. Hazledine / Journal of Air Transport Management 17 (2011) 130e135
and flexibility in terms of permitted itinerary changes. Also, tickets purchased in the lowest (“Smart Saver”) fare class did not earn frequent flier points, nor status points; Each fare type on each route had four set price points, with no overlap across the fare types, so that in essence there was a ladder of twelve prices which could be used for each flight (supplemented by occasional deep discount sales); All tickets one-way7; Everything, including FFP bookings directly accessible to consumers on the airline’s transparent, easy to use website8 While some features of the Express system can be seen as decreasing the airline’s vulnerability to LCC competition (stripping out onboard costs; abandoning return fares, promoting direct B-2-C selling from the website), the initiative actually went further than that, to capitalise on Air New Zealand’s position as the national carrier with a near-complete route structure, covering every region and most towns.9 The non-linearities in national TV advertising and in producing a national website were exploited just about every person seeing the advertisements or looking up the website was a potential customer. The slogan used to market the new fare system was “Being There”, with the message that air travel was now so cheap and convenient and easy to choose and purchase that New Zealanders of all types from all regions should thing nothing of hopping on a plane to “be there” at the grandchild’s christening or the friend’s 21st birthday party. The Express System was promptly matched by Air New Zealand’s domestic competitor Qantas, and soon extended, with some changes because of the longer journeys e e.g. meal service retained along with the business class cabin, to the trans-Tasman routes. In Air New Zealand’s 2003 Annual Report it was claimed that: Simplicity is the essence of affordable travel. Fare structures were extensively simplified and prices reduced by an average of 20% and up to 50%. The everyday low fares stimulated traffic by 22%. capacity increased by 10%. The increase in capacity was particularly interesting because it was achieved at minimal direct cost, by eliminating business class seats and by achieving faster turn-around times with no meals and beverages to load and unload. The airline reported that it had decreased the fares paid by its business travellers10, as well as the leisure market. Overall, the innovation can be seen as an unusual
7 Air NZ, perhaps uniquely amongst the world’s legacy carriers, had long offered discounted one-way fares. I have a 1993 printed timetable in which a suite of five different economy fares is set out for every route. The cheapest of these, “Super Thrifty”, offered a 45% discount on the regular economy fare, on selected [off-peak, presumably] flights, with just a 7-day Advance Purchase requirement. The reason for this may be geography and demographics. NZ is a thousand miles long, and not very wide. Many tourists like to fly one-way and drive the other, in rental cars or camper vans. Young NZers often attend colleges and universities away from home, which also creates a strong demand for cheap one-way fares. 8 Following the Express Fare system focused on in this paper, Air New Zealand was possibly the first airline to simplify its FFP by making reward points fully convertible to “Airpoints dollars” which can be used to purchase any ticket from the website, as an alternative to a cash (direct debit or credit card) purchase. It also implemented a total remodelling of its long-haul product, including a leading-edge business class lie-flat “pod” and a new “premium economy” class, and has been quite active in repositioning its international route structure. 9 Air New Zealand’s domestic fleet ranges from B737e300s for the main trunk and tourist routes, down through ATR72 and Bombardier Q300 turboprops, to 19seat Beechcraft 1900D aircraft. The smaller planes allow the airline to offer a convenient service (multiple daily departures) on relatively “thin” routes. LCCs, with their single-type jet fleets, cannot match this. 10 “More importantly, we’re tracking airfares purchased by our business travellers and are pleased to see savings of around 21% being made in comparison to what they would have paid for the same journeys before Express” (Air NZ internal enewsletter, July 1, 2003).
and unusually successful attempt by a mature business to increase profitability the hard way by moving down its demand curve, as well as shifting out that curve by reducing transaction costs and boosting the idea of easy flying. The price point fare system enables the airline to change price without changing prices, as Table 1 demonstrates. This shows the range of lowest prices offered on the Air NZ website for flights on eight domestic routes, observed on eight successive Wednesdays from November 2004 to early January 2005 (Hazledine, 2010). There were 69 different flight numbers, and each flight/date was observed weekly nine times, from eight weeks to one day before take-off. Thus there were close to 5000 potential observations of fares (actually less than that because some flights sell out and are removed from the website).11 Table 1 reveals that the pricing structure was stable over the eight week period, and that just 83 different fares (of the 96 price points available) were used, but this still enabled the airline to effect a considerable amount of price dispersion. Some of this was intertemporal price discrimination, with lowest fares tending to increase in the last two weeks, which is when business and other urgent travellers come into the market. Some of the dispersion was yield management, with flights known to be popular for business travellers (in particular early morning AucklandeWellington flights) being priced higher than off-peak flights even two months in advance to keep the seats available for the high yielding customers expected later. Air New Zealand’s Express Fare system remains in use in 2009, though with gradual modifications. Since December 2008 trips ticketed on the cheapest Smart Saver fares count for status points, though still not FFPs. It is now possible to change Smart Saver tickets before travel, for a fee. Perhaps because the middle fare class was being crowded by these relaxations of Smart Saver conditions, since October 2009 the two higher fare classes have been amalgamated into one (“Flexi Plus”), which permits free flight changes and cancellations, thus focusing on the key attraction for high willingness to pay business travellers flexibility. Perhaps because the statutorily required number of flight attendants (three on B737s) don’t now have much work to do, a bar service has been reintroduced for weekday flights leaving between 5 and 7pm.12 A popular innovation was “grabaseat”: sometime after midnight each day the front page of the website is loaded with a link to a set and known number of seats on a variety of routes, usually around four to six, for some future travel period over one week, at extremely low prices, as low as $1. In an informal seminar at the University of Auckland in early 2007, an Air New Zealand pricing manager explained that the ‘ludicrously low fares’ available on grabaseat were offered to ‘maintain [consumer] low-fare perception’ of the airline, and to ‘drive repeat internet traffic and travel habits’. The manager also revealed that Air NZ’s domestic market was perhaps the most profitable segment of the airline’s overall route structure, despite competition from Qantas and, since late 2007, Pacific Blue, in what is ‘possibly the easiest place in the world to set up an airline’. The profitability situation on the trans-Tasman market is not so rosy, with plentiful capacity keeping pressure on prices e a situation that, from the airlines’ point of view, will have been exacerbated
11 That is, each observation is of the lowest fare available at the time. Of course when that fare was a “Smart Saver” fare tickets could also be purchased in the “Flexi Saver” and “Flexi” fare classes. Thus, when, say, the Flexi fare was observed (usually close to travel date), it means that the lower fare classes or “buckets” had by then been removed from the market. 12 Although flights leaving from locations where it would be inconvenient to load a bar are scheduled to depart at 455pm rather than the 500pm which turns out to be their actual take-off time.
T. Hazledine / Journal of Air Transport Management 17 (2011) 130e135
133
Table 1 Prices observed on Air New Zealand, NovembereDecember 2004. Route
Distance (Kms)
Daily number of flights
SuperSaver fare class
AucklandeDunedin AucklandeNapier AucklandeQueenstown AucklandeWanaka AucklandeWellington ChristchurcheQueenstown ChristchurcheWanaka WellingtoneChristchurch
1062 328 1023 1000 480 347 294 303
12 11 6 1 19 6 1 13
109
129
128
148
90 84 86 70
100 94
139 86 168 155 110 104
169 96 188 185 120 114
80
90
100
since the February 2009 introduction of Emirates’ new A380 superjumbo on its daily AucklandeSydney flight. Nevertheless, Air New Zealand’s relative position seems to be better than the other carriers, in terms of yield per passenger. Table 2 is digested from Hazledine (2008) and shows the coefficients of airline dummy variables in lowest fare regressions (i.e.,using observations like those summarised for Air NZ’s domestic routes on Table 1) controlling for costs, market structure and demand shift factors which are not here shown. The database includes the 2004e05 domestic NZ observations, and also trans-Tasman flights observed in mid 2005 and mid 2006. The coefficients on the airline dummies show the airlines’ prices relative to Air New Zealand’s price, other factors held constant. The three pricing variables are, first, a (weighted) average of all observed lowest fares offered for each flight; second, the lowest fare available eight weeks out, which usually but not always is the lowest fare of all, and, third, the lowest fare available the day before the flight, which for most but not quite all flights is the highest lowest fare observed. The results are quite striking. Overall in these markets Air New Zealand is able to charge a substantial fare premium over its rivals e fares more than ten percent higher than its legacy airline competitor Qantas, and around twenty five percent above either the LCC Pacific Blue or the 5th Freedom carrier Emirates.13 The price differentials are much smaller eight weeks before the flight, when leisure travellers dominate the market. It is in the more lucrative late-purchasing business segment of the market that Air New Zealand is able to pull away from its rivals including Qantas in terms of its yields, being able to sustain price differentials of close to 30% over both Pacific Blue and Emirates. There is little evidence here to support the proposition that LCC competition, granted its overall impact on fare structures, has made legacy carrier pricing practices, and in particular, price discrimination, obsolete. 4. The legacy response II: Air Canada American Airlines’ Saturday Night Stayover restriction, brilliant though it was, came at a cost. It amounts to “spoiling” the product, because it must deter leisure travellers who either do not wish to make a return trip (they want a one-way or perhaps an open-jaw itinerary), or who do not wish to stay away for a Saturday night. Selling cheap one-way tickets, as does Air New Zealand, of course eliminates that element of spoiling of the product, which is no doubt part of its success in increasing sales. The Express Fare system, however, still retains a market-spoiling element from the old legacy pricing regime. Note from Table 1 that there is no overlap between the fare class price points. And recall
13
FlexiSaver fare class
Note that Pacific Blue and Emirates each only competes on a subset of transTasman routes. Also note that the Qantas differential is in real terms even larger than the coefficient implies, because this airline gives FFPs in all its fare classes, and serves hot food and beverages on its domestic NZ flights.
199 106 218 215 150 144 136 120
219 126 248 245 170 164 156 140
239 146 278 275 190 184 186 150
Fully Flexi fare class 279 176 308 315 200
299 206 338
160
180
230 229
399 226 488 375 280 254 242 220
479 246 608 485 330 276 276 260
559 306 708 360 314 290
that the use of the system to implement inter-temporal price discrimination basically involves progressively removing lower fare “buckets” from the market, until, in the last week or so before flight date, only “Flexi” fares may be available. That is, the higher fares come with additional features bundled in. But many leisure travellers, and even some business travellers, may place little or no value on such fripperies as FFP points, status points, and even flexibility to make itinerary changes. Then, if there is any cost in supplying these features, it may be inefficient to bundle them rather than offer a no-frills product as an option. Perhaps in response to these considerations, Air Canada has taken the Air NZ innovations another step e perhaps not a step further, but ultimately in a strikingly different direction. Superficially, Air Canada’s domestic fare offerings look rather like Air New Zealand’s. There are three economy cabin fare classes14 e Tango, Tango Plus, Latitude-with more features and higher prices in the higher classes, and tickets are all one-way, of course. But whereas in the Air NZ’s case one gets the impression that the fare classes really mostly serve as a convenient way to divide up the price points into digestible sets e e.g. three sets of four points rather than a long ladder of twelve points-for Air Canada, the fare classes have become important products in themselves. Insights into the airline’s strategy can be gleaned from various speeches by senior executives preserved on their website. Thus, in May 2006, it was stated that the goal was to have at least five “options” per fare “product” e that is, creating distinct groupings of characteristics, each group developing its own integrity as a recognised product in the market place.15 So, the “products” each have their own slogans on the website, along with suggestive images of the target customers: a young blond woman for Tango (“Our best value”); a laughing black man for Tango Plus (“Get up and go!”); a cool white dude for Latitude (“A perfect fit”); and a handsome mature gentleman for Executive Class (“maximum comfort and freedom”). 16 A basic requirement for any product to have integrity in the mind of consumers is that it be regularly and dependably available, and here we see the first big difference from Air NZ. Whereas it is quite unusual for the latter airline to still offer its lowest Smart Saver fare the day before travel, in the database David Gillen and I have assembled on Air Canada, we find that in about 90% of flights the Tango Plus, if not the Tango, fare is still available the day before, on domestic routes. Now, this fact would not be so interesting if the differences in prices of the fare products were small, but they are not. The premium for Tango Plus over Tango is actually quite low e $35 is
14 In 2006, Air Canada had a fourth economy fare class e Latitude Plus e but this has subsequently been eliminated. 15 http://www.aircanada.com/en/about/investor/documents/sctac-may06.pdf. 16 http://www.aircanada.com/en/about/media/presentations/documents/NBF_ March_28.pdf.
134
T. Hazledine / Journal of Air Transport Management 17 (2011) 130e135
Table 2 Airline dummies in airfare regressions. Sample: 1963 domestic NZ and trans-Tasman flights observed 2004e06
Weighted average of nine lowest price observations
Lowest price observed eight weeks before flight date
Lowest price observed one day before flight date
Dependent variable Qantas PBlue Emirate
Log(Pwavk) 0.117a 0.239a 0.225a
Log(P8k) 0.087a 0.102 0.023
Log(P0k) 0.182a 0.316a 0.314a
a
Significant at the 1% level.
common and the benefits quite substantial: more status miles, more FFP points, access to the Lounges for $30; eligibility to use upgrade certificates to business class, subject to availability. Here the airline is trying to encourage “buy-up”, and in its website presentations it has boasted achieving in 2006 a “48% buy-up”, which presumably means that nearly one third of the passengers purchasing one of the low-fare products paid to get Tango Plus. But the truly striking, even, remarkable feature of Air Canada’s system is that the price differential between Tango Plus and Latitude is huge: around $200, on average, as we will see. True, Air New Zealand’s fare class price differentials are of similar size, but the point is that, as noted above, last minute travellers generally cannot choose between both products e the Smart Saver fares have been taken off the market. So Air Canada, to preserve the integrity of its fare products, is offering seats on the same flight the day before travel at substantially different prices. But why would anyone purchase the more expensive ticket? Well, do they indeed purchase Latitude fares? Again, from the presentations stored on the website, we find the statement: “Tango only accounts for 45% of domestic sales in Quarter 4 [2006]”, which with the 48% buy-up to Tango Plus, implies that around one third of tickets sold were either in business class or were Latitude fares. Now, from the demand side, it is quite hard to match the price differential charged for Latitude over Tango Plus with the additional features obtained. These are: no $50 change fee, complimentary snack, more FFP miles, plus better chances of being able to use an upgrade certificate to the business class cabin. What this probably bears witness to is the agency problem created when business and government travellers don’t have to personally bear the cost of their tickets. Be this as it may, there are clearly large numbers of domestic Air Canada travellers who are willing to pay or to get someone else to pay for various perks and frills which Air Canada can supply and its LCC domestic rival WestJet cannot. Furthermore, looking now at the supply side, it seems that the profit margins on these frills must be substantial. Indeed, the Latitude fare is a product created, almost literally, out of nothing (apart from the cost of flying the customer, which is common to all the fares). It is almost costless to supply these frills, given the non-linearities of operating a full service national network carrier. There are substantial fixed costs and low variable costs in setting up a business class cabin and service, in operating a network of lounges, in setting up and marketing the FFP, in joining the Star Alliance, as well as in operating a full regional network and marketing campaign. For example, the smallest business class cabin operated by Air Canada has nine seats (on its Embraer 175 and 190 aircraft), because of fixed costs. When some of these seats have not been sold to full-fare paying customers, it is virtually costless at the margin to allow high-status or Latitude fare economy customers to upgrade. Indeed, the airline may thereby free up economy seats which can be sold profitably to standby customers. Air Canada’s opaque (in comparison to Air NZ) FFP rewards redemption processes appear to
be managed (as even high-status Aeroplan members know) so as to eliminate opportunity costs of making seats available. And, of course, the non-linearities of the reward schedules effectively lock many fairly frequent Canadian fliers into the Air Canada plan. Access to Maple Leaf Lounges at $25 or $30 probably easily covers the marginal costs of the food and beverages consumed within. So, it seems that Air Canada has come up with a pricing and marketing strategy which allows it to generate quite large revenues from its more-or-less locked in high willingness to pay customers, whilst still covering itself against LCC competition from WestJet by keeping a regular and quite dependable supply of budget fares in the market at all times.17 Table 3 shows how this works out. Here we examined Air Canada’s fares charged in the context of three different market structures: domestic Canada monopoly; domestic Canada duopoly with WestJet, and trans-border routes on which it competes with usually three or four or more US legacy carriers (some serving the routes through code shares, usually with global Alliance partners). To standardise for the very different route lengths, which of course are reflected in fare structures, all the fares are converted to 1400 km journey equivalent, using the coefficient on the log of route distance in an econometric pricing model.18 This is the average distance flown in our sample of routes, which includes 380 Air Canada flight numbers on 46 domestic and 27 trans-border routes, observed weekly from three weeks before to one day before take-off, for flights departing on three consecutive Wednesdays in October 2007. 17 of the domestic routes were then Air Canada monopolies; 29 were shared with WestJet. The first two columns on Table 3 show the (average) Tango Plus and Latitude fares offered 22 days before the flight date; the third and fourth columns show these fares one day before the flight, and columns five and six are the price of the “frills” e i.e., the premium paid for a Latitude ticket over the price of a Tango Plus ticket e 22 and 1 day before the flight. Look first at Air Canada’s behaviour on its nineteen domestic monopoly routes. We see that there was quite a lot of intertemporal price discrimination, with Tango Plus fares moving up 37% over the three weeks, on average.19 However, there is not a lot of change in the frills price, which just slips from $213 to $197. This may not mean a lot, because most business travellers don’t appear in the market until a week or ten days before the flight, so that the Latitude fare 22 days out is probably not heavily sold. Now look at the duopoly routes. It is strikingly clear that competition makes a difference to Air Canada’s pricing e its Tango Plus fares are just about two thirds as high as on the monopoly routes. But is it “low-cost” competition that matters? Possibly not. The frills markup is a bit lower in dollars per km; a bit higher as a percentage, in the duopoly markets. That is, the pricing game that Air Canada is playing with its “fare products” does not seem to be
17 Although the Tango or Tango Plus fare was not available day before for 10% of Air Canada’s flights in our database, it was almost invariably possible to purchase a cheap ticket on some e usually mid morning or afternoon or red-eye e flight on the same route the day before, and it seems that that fare was typically the same as the lowest available WestJet fare on the route. Thus, the proud quoting by Air Canada executives of their rival WestJet’s CEO Clive Beddoe’s statement: “Air Canada matches us dollar for dollar on every single fare” is not exactly false, though it could be misleading [http://www.aircanada.com/en/about/media/presentations/ documents/NBF_March_28.pdf]. 18 The coefficient on route distance was 0.75 in a model with price per kilometre dependent. That is, other things equal, the actual fare (ticket price) increases by 25% for each doubling of route distance. 19 Although it is not demonstrated in the average figures of Table 3, it is quite common for the day 1 Tango Plus fare to exceed the Day 22 Latitude fare for a particular flight. This is another difference with Air New Zealand’s Express system, where the price points always move up monotonically through the fare classes.
T. Hazledine / Journal of Air Transport Management 17 (2011) 130e135
135
Table 3 Air Canada’s Strategic Product Enhancement. Average fares and prices of frills, adjusted to 1400 km journey ($)
Canada monopoly routes Canada duopoly routes Trans-border routes All routes
Tango þ fare, 22 days before take-off
Latitude fare, 22 days before take-off
Tango þ fare, 1 day before take-off
Latitude fare, 1 day before take-off
Price of frills 22 days before take-off
Price of frills 1 day before take-off
473 302 309 336
686 480 517 529
648 430 767 564
846 619 858 722
213 178 209 193
197 189 91 158
Note: Data from Gillen and Hazledine (forthcoming), prices observed on Air Canada website for 46 domestic Canadian and 27 trans-border routes; 380 flight numbers; each flight observed four time from 22 days to 1 day before take-off, on three Wednesdays in October, 2007.
much affected by the presence or otherwise of its LCC domestic rival, WestJet. The contrast with the trans-border routes is quite dramatic. With the exception of a couple of Florida destinations (Orlando and Tampa) to which WestJet flies, Air Canada is now getting its competition from a bevy of US-based legacy carriers. These are airlines which have not changed a great deal in their marketing practices. They still offer discounts on return tickets and on Saturday Night Stayovers20. Their one-way ticket offerings are quite rudimentary e you can buy a refundable or non-refundable ticket. So, with the exception of the Florida routes, Air Canada transborder does not offer Tango fares. And, although the Tango Plus/ Latitude difference 22 days out is in line with the domestic differences, when we get to the business end of the market close to the flight date, the differential is sharply compressed. Basically, Air Canada matches the refundable/non-refundable fare differences charged by its US rivals, these being on some routes as low as $20. The fact that Air Canada can do this, and presumably not lose money, is evidence on just how cheap the Latitude frills must be to supply. 5. Conclusion These case studies document two legacy airlines’ innovative and spirited responses to the low-cost carrier challenge, and demonstrate, at the very least, that even well-run LCCs are not a force majeure, sweeping aside all resistance. The simplicity which is the strength of the low-cost business model is also a potential source of weakness, that can be exploited by a legacy carrier determined to extract full value from its unique fixed assets. There may, however, be some doubts about the applicability of the Air Canada and Air New Zealand strategies to the situation of legacy carriers in larger markets such as the US and Europe. These two airlines have the advantages of national carriers, which include a myriad of formal and informal business linkages within their home market; complete regional coverage of this, and the capacity to operate on long-haul routes where LCCs cannot follow. Thus, it is quite striking that Air Canada is apparently unable to extract the same markup on its premium fare class on trans-border routes, where its competition is other legacy carriers, as it earns domestically in competition only with WestJet. The practices of the US legacy
20 Puller et al. (2009) find that round-trip and SNS tickets come with 11% and 13% discounts, in the US.
carriers, however, which still impose return-journey restrictions on their low-fare tickets, are coming to seem rather antediluvian, and it is just possible that the examples of two “foreign” carriers’ quite successful responses to LCC competition might give them the nerve to try something innovative themselves.
Acknowledgements Without implication, the author thanks two referees and participants in the 2009 GARS Workshop, the 12th Hamburg Aviation Conference and the Sullberg Society 3rd Meeting for comments and insights, especially Nicole Adler, Achim Czerny, Hans-Martin Niemeier, Mike Tretheway, Martin Dresner, David Gillen, Ken Button, Anming Zhang and David Starkie. Some of the material used in this paper is drawn from research in progress joint with David Gillen.
References Berry, S., Jia, P., 2009. Tracing the Woes: An Empirical Analysis of the Airline Industry. Annual Conference. American Economics Association, San Francisco. Button, K.J., 2004. Does the theory of the ‘core’ explain why airlines fail to cover their long-run costs of capital? Journal of Air Transport Management 9, 5e14. Dana, J.D., Orlov, E., 2009. Internet Penetration and Capacity Utilization in the US Airline Industry. Annual Conference. American Economics Association, San Francisco. January. Dresner, M., Lin, J.-S., Windle, R., 1996. The impact of low-cost carriers on airport and route competition. Journal of Transport Economics and Policy 40, 473e571. Goolsbee, A., Syverson, C., 2008. How do incumbents respond to the threat of entry? Evidence from the major airlines. Quarterly Journal of Economics 123, 1611e1633. November. Gillen, D., Gados, A., 2008. Airlines Within Airlines: Assessing The Vulnerabilities of Mixing Business Models. Research in Transportation Economics 24, 25e35. Gillen, D., Hazledine, T., forthcoming. The New Pricing in Canadian Air Travel Markets: Implications for Competition and Antitrust. In: James Peodes (Ed.), Advances in Airline Economics, Vol. 3, Emerald Books. Working Paper. Centre for Transportation Studies, University of British Columbia. Hazledine, T., 2008. Modern Airfare Price Discrimination: Theory and Evidence. University of Auckland. Hazledine, T., Pricing., 2010. competition and policy in Australasian air travel markets. Journal of Transport Economics and Policy 44, 37e58. Puller, S.L., Sengupta, A., Wiggins, S.N., 2009. Testing Theories of Scarcity Pricing and Price Dispersion in the Airline Industry. Annual Conference. American Economics Association, San Francisco (January). Tretheway, M., 2004. Distortions of airline revenues: why the network airline business model is broken. Journal of Air Transport Management 10, 3e14.
Contents lists available at ScienceDirect
Journal of Air Transport Management journal homepage: www.elsevier.com/locate/jairtraman
Legacy carriers fight back: Pricing and product differentiation in modern airline marketing Tim Hazledine Department of Economics, The University of Auckland, New Zealand
a b s t r a c t Keywords: Airline pricing Low-cost airlines Legacy airlines
The adoption of the low-cost carrier business model has applied competitive pressure on established network or “legacy” carriers, by offering fares at prices that legacy carriers find it difficult to match and still cover their fixed costs. This paper reports how two medium-sized national airlines-Air New Zealand and Air Canada-have coped with the low-cost threat by, in effect, turning their fixed costs into profit centres. Features such as full regional networks, long-haul connections, frequent flyer programs, membership in global alliances, lounges and business class cabins can be bundled into products which can be marketed and sold profitably to business and even some leisure travellers, and which cannot be easily replicated by low-cost carriers. Although not panaceas, the innovations of Air New Zealand and Air Canada to the competition they face in their domestic and trans-border markets demonstrate the possibility of an effective legacy carrier response to the low-cost carrier business model. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction This paper examines how two full service national network carriers e Air New Zealand and Air Canada-have responded to two of the major challenges facing airlines in the 21st century: namely, the continued competitive threat from the low-cost carrier (LCC) business model, and the potentialities of the internet as an instrument for marketing and transacting air carrier services. The analytical context is the “empty core” hypothesis (Button, 2004), to the effect that the large fixed costs incurred by “legacy” carriers make it difficult-perhaps impossible-for them to set prices adequate to cover all their costs without those prices being well above marginal variable costs, and therefore vulnerable to predatory marginal cost pricing by rivals, with this unstable situation presumably exacerbated when the rivals are low-cost carriers without large fixed costs of their own to cover because of their flexible, route profitability-based business model. Some legacy carriers have responded to the LCC threat by stripping out their own fixed costs e either by in effect turning themselves into low-cost carriers on short-and medium haul routes (e.g. Aer Lingus), or by setting up LCC subsidiaries and deploying them as a “fighting-brand” on the leisure-oriented routes most susceptible to LCC competition (e.g. Qantas with its subsidiary Jetstar).1
E-mail address: [email protected]. Air Canada (“Tango”) and Air New Zealand (“Freedom Air”) have both experimented with fighting-brand LCC subsidiaries, neither of which are still in operation. 1
0969-6997/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jairtraman.2010.10.008
In contrast, the strategies of Air New Zealand and Air Canada share a quite different response to the LCC business model. Instead of stripping out fixed costs, they turn them to their advantage. The basic idea of this paper is that legacy carriers’ fixed costs can actually deliver market value; that this value can be enhanced by shrewd and innovative pricing and marketing practices, and that such practices cannot easily be replicated by LCCs and so can deliver a source of competitive advantage to the legacy carriers which may compensate for the LCCs’ lower variable costs. These valuable “fixed” elements in Air New Zealand and Air Canada’s product portfolios include their status as national carriers, the density and coverage of their domestic and international route structures, the penetration of their websites, their membership of global alliances, their FFPs, their lounges and other benefits for high-status travellers, and their business class cabins. 2. Background: market penetration of LCCs Although the market penetration of low-cost carriers did increase quite rapidly over the past two decades, there may now be signs emerging of some sort of equilibrium, in which the overall market share of LCCs has stuck at numbers in the 20e40 percent range, depending on market structure. In the US, the LCC share of the domestic market, in terms of number of passengers increased from 4% in 1990 to 20% in 1999, and thence to 33% in 2008, but the share of the six largest legacy carriers was 63.3% in 1997 and 63.1% in 2005, so that it may be that much of increase in LCC penetration is due to the adoption of the LCC business model by most start-ups
T. Hazledine / Journal of Air Transport Management 17 (2011) 130e135
and other small airlines, rather than to a push-back of the positions of the major legacy suppliers.2 Note too that legacy carriers fly longer routes, on average, and so their RPM (revenue passenger miles) share is substantially larger, and their overall revenue share would be larger still, since their fares tend to be higher than those of the LCCs.3 Now, it could be argued that holding onto to market share is not the point e what matters is the cost of so doing. A succession of studies, e.g. Dresner et al. (1996) and most recently Goolsbee and Syverson (2008), have documented the dramatic impact on legacy carrier prices on routes impacted by actual or even just increased probability of entry by LCCs; in particular Southwest. Market share defended by slashing prices when costs are not slashed will impact profitability, and may not be sustainable, as Tretheway (2004) has argued. Nevertheless, the Big-6 (now Big-5) US legacy carriers have chosen to respond to LCC competition with lower prices rather than just by ceding market share, and we should not presume that they don’t have their reasons for doing so. As for profitability e well, the airline industry in North America has somehow survived for decades with long-run rates of return apparently well below what would be consider acceptable in most other sectors. Aviation is a glamorous business, and it may be that the “normal” rate of return on capital is lower in this sector, as it seems to be in other inherently attractive “lifestyle” activities, such as movie production, wineries and even farming in general. The recent spate of Chapter 11 episodes besetting the legacy carriers may in part reflect cash flow problems generated by sharper competition in the marketplace, but probably also has something to do with what could be called the demographic implications of their situation, whereby the incumbents are stuck with a literal legacy of pension and health plan obligations incurred in easier days, as well as a relatively highwage unionised wage structure.4 Berry and Jia (2009) carry out an extremely sophisticated analysis of air travel demand, supply, and profitability in 1999 and 2006, and find that only a quite small proportion e around 10%-of the observed reduction in legacy carriers’ profits between the two years were apparently directly due to the expansion of LCCs over that period, with major factors being greater price sensitivity of customers and a systematic change in preferences towards direct (not connecting) flights. The higher demand price elasticities have been informally linked to greater consumer use of internet ticketing: Dana and Orlov (2009) provide direct evidence that increases in internet penetration can explain the dramatic increase in load factors observed in US markets from just 62%, on average, in 1993 to 80% in 2007. Smaller domestic markets support fewer airlines. In Canada, the near-monopoly created by the folding of the failed CAIL into Air Canada has been challenged successfully by the LCC WestJet, which was launched in 1996 and now takes about 40% of the domestic market as well as some of the trans-border business. In Australia, the failure of Qantas’s long-standing legacy rival Ansett, in 2001, permitted the recently arrived LCC Virgin Blue to expand quickly to
2
These data are taken from tables analysing US Department of Transportation DB1A data, found on www.darinlee.net. The market share of the Big-6 (AA, Continental, Delta, NWA, United, US Airways) was boosted by the absorption of TWA by American in 2002 and of America West by US Airways after 2005. 3 BTS data reported on their RITA website show the six largest legacy carriers holding 47% passenger share and a 58.5% RPM share, in the months up to October 2008, with the same statistics for the three largest LCCs (Southwest, AirTran, JetBlue) being 22.6 and 20.3%,. The discrepancy between these numbers and those reported above may be because the latter uses US Department of Transportation data which do not include all the smallest airlines. 4 A referee has suggested that these legacies may create barriers to down-sizing, which in turn may help explain why the US legacy carriers have fought so hard to retain market share.
131
take around one third of the domestic market e a figure which appears to be acceptable to both airlines and which has been quite stable. In New Zealand, although the inhabitants apparently consume more air miles per capita than in any other country, the size of the market (over four million residents plus over two million annual tourists) has historically been able to support only one profitable carrier (the national incumbent Air New Zealand), despite which first Ansett, then Qantas and now Pacific (Virgin) Blue as well as Qantas have taken advantage of the totally Open Skies regulatory environment to offer domestic services on main trunk and important tourist routes. There are no official data, but Air New Zealand appears to offer more than two thirds of the seats on the routes on which it faces competition, and has a monopoly of the other, regional, routes.5 On the trans-Tasman flights of 3e4 h duration that link the main cities of Australia and New Zealand, the legacy incumbents Air New Zealand and Qantas have since 2002e03 faced competition from Pacific Blue and from Emirates Airlines (making use of 5th Freedom rights), but seem able to hold on to market shares of between 70% and 80% in most markets. 3. The LCC competitive threat: Air New Zealand’s response Although the most obvious competitive challenge posed by the more successful LCCs was the substantially lower price tags on their product, it was the nature of the product itself that may in the longrun have posed the biggest threat to the legacy business model. Indeed, Tretheway judges that: Perhaps the most important impact of the LCC business model on [legacy carriers] has been the introduction of low one-way fares. This has undermined the price discrimination ability of the [legacy carriers], and is the most important pricing development in [the] past 25 years. (2004, p. 5) If the (re)introduction of “low” one-way fares (i.e., not full-fare, unrestricted products aimed at business travellers) is the most important development in a quarter century, this reflects the fact that it challenged what arguably was the most important pricing innovation twenty five years ago e namely, the introduction in the early 1980s by American Airlines of the advance purchase discounted return fare with its key Saturday night stayover (SNS) condition. This brutally simple marketing ploy was of course a clever way of getting high willingness to pay business travellers to self select themselves away from low-WTP leisure customers, with the latter willing and often even keen to secure their ticket well in advance, and undeterred by the SNS condition because most leisure itineraries include a weekend away in any case. The first legacy carrier to introduce a fully worked through response to the LCC business model6 appears to have been Air New Zealand, who in December 2002 rolled out their new ‘Express Fare’ system. The Express system had the following features: The product was pared back, with business class seating taken out, and hot meals and bar service removed, on domestic routes; The previous large number of different fare types or buckets was simplified to three, differing mainly in their refundability
5 Qantas has made no secret of its difficulties in earning profits from its domestic NZ operations. Most recently, the CEO of the airline was reported to have said ‘Qantas came close last year [2008] to exiting the New Zealand domestic market because it was haemorrhaging money’ [See http://www.stuff.co.nz/business/ industries/2924832/Airlines-bleed-in-Tasman-route-scrap]. 6 For a good taxonomy of the differences between legacy and LCC business models, see Gillen and Gados (2008).
132
T. Hazledine / Journal of Air Transport Management 17 (2011) 130e135
and flexibility in terms of permitted itinerary changes. Also, tickets purchased in the lowest (“Smart Saver”) fare class did not earn frequent flier points, nor status points; Each fare type on each route had four set price points, with no overlap across the fare types, so that in essence there was a ladder of twelve prices which could be used for each flight (supplemented by occasional deep discount sales); All tickets one-way7; Everything, including FFP bookings directly accessible to consumers on the airline’s transparent, easy to use website8 While some features of the Express system can be seen as decreasing the airline’s vulnerability to LCC competition (stripping out onboard costs; abandoning return fares, promoting direct B-2-C selling from the website), the initiative actually went further than that, to capitalise on Air New Zealand’s position as the national carrier with a near-complete route structure, covering every region and most towns.9 The non-linearities in national TV advertising and in producing a national website were exploited just about every person seeing the advertisements or looking up the website was a potential customer. The slogan used to market the new fare system was “Being There”, with the message that air travel was now so cheap and convenient and easy to choose and purchase that New Zealanders of all types from all regions should thing nothing of hopping on a plane to “be there” at the grandchild’s christening or the friend’s 21st birthday party. The Express System was promptly matched by Air New Zealand’s domestic competitor Qantas, and soon extended, with some changes because of the longer journeys e e.g. meal service retained along with the business class cabin, to the trans-Tasman routes. In Air New Zealand’s 2003 Annual Report it was claimed that: Simplicity is the essence of affordable travel. Fare structures were extensively simplified and prices reduced by an average of 20% and up to 50%. The everyday low fares stimulated traffic by 22%. capacity increased by 10%. The increase in capacity was particularly interesting because it was achieved at minimal direct cost, by eliminating business class seats and by achieving faster turn-around times with no meals and beverages to load and unload. The airline reported that it had decreased the fares paid by its business travellers10, as well as the leisure market. Overall, the innovation can be seen as an unusual
7 Air NZ, perhaps uniquely amongst the world’s legacy carriers, had long offered discounted one-way fares. I have a 1993 printed timetable in which a suite of five different economy fares is set out for every route. The cheapest of these, “Super Thrifty”, offered a 45% discount on the regular economy fare, on selected [off-peak, presumably] flights, with just a 7-day Advance Purchase requirement. The reason for this may be geography and demographics. NZ is a thousand miles long, and not very wide. Many tourists like to fly one-way and drive the other, in rental cars or camper vans. Young NZers often attend colleges and universities away from home, which also creates a strong demand for cheap one-way fares. 8 Following the Express Fare system focused on in this paper, Air New Zealand was possibly the first airline to simplify its FFP by making reward points fully convertible to “Airpoints dollars” which can be used to purchase any ticket from the website, as an alternative to a cash (direct debit or credit card) purchase. It also implemented a total remodelling of its long-haul product, including a leading-edge business class lie-flat “pod” and a new “premium economy” class, and has been quite active in repositioning its international route structure. 9 Air New Zealand’s domestic fleet ranges from B737e300s for the main trunk and tourist routes, down through ATR72 and Bombardier Q300 turboprops, to 19seat Beechcraft 1900D aircraft. The smaller planes allow the airline to offer a convenient service (multiple daily departures) on relatively “thin” routes. LCCs, with their single-type jet fleets, cannot match this. 10 “More importantly, we’re tracking airfares purchased by our business travellers and are pleased to see savings of around 21% being made in comparison to what they would have paid for the same journeys before Express” (Air NZ internal enewsletter, July 1, 2003).
and unusually successful attempt by a mature business to increase profitability the hard way by moving down its demand curve, as well as shifting out that curve by reducing transaction costs and boosting the idea of easy flying. The price point fare system enables the airline to change price without changing prices, as Table 1 demonstrates. This shows the range of lowest prices offered on the Air NZ website for flights on eight domestic routes, observed on eight successive Wednesdays from November 2004 to early January 2005 (Hazledine, 2010). There were 69 different flight numbers, and each flight/date was observed weekly nine times, from eight weeks to one day before take-off. Thus there were close to 5000 potential observations of fares (actually less than that because some flights sell out and are removed from the website).11 Table 1 reveals that the pricing structure was stable over the eight week period, and that just 83 different fares (of the 96 price points available) were used, but this still enabled the airline to effect a considerable amount of price dispersion. Some of this was intertemporal price discrimination, with lowest fares tending to increase in the last two weeks, which is when business and other urgent travellers come into the market. Some of the dispersion was yield management, with flights known to be popular for business travellers (in particular early morning AucklandeWellington flights) being priced higher than off-peak flights even two months in advance to keep the seats available for the high yielding customers expected later. Air New Zealand’s Express Fare system remains in use in 2009, though with gradual modifications. Since December 2008 trips ticketed on the cheapest Smart Saver fares count for status points, though still not FFPs. It is now possible to change Smart Saver tickets before travel, for a fee. Perhaps because the middle fare class was being crowded by these relaxations of Smart Saver conditions, since October 2009 the two higher fare classes have been amalgamated into one (“Flexi Plus”), which permits free flight changes and cancellations, thus focusing on the key attraction for high willingness to pay business travellers flexibility. Perhaps because the statutorily required number of flight attendants (three on B737s) don’t now have much work to do, a bar service has been reintroduced for weekday flights leaving between 5 and 7pm.12 A popular innovation was “grabaseat”: sometime after midnight each day the front page of the website is loaded with a link to a set and known number of seats on a variety of routes, usually around four to six, for some future travel period over one week, at extremely low prices, as low as $1. In an informal seminar at the University of Auckland in early 2007, an Air New Zealand pricing manager explained that the ‘ludicrously low fares’ available on grabaseat were offered to ‘maintain [consumer] low-fare perception’ of the airline, and to ‘drive repeat internet traffic and travel habits’. The manager also revealed that Air NZ’s domestic market was perhaps the most profitable segment of the airline’s overall route structure, despite competition from Qantas and, since late 2007, Pacific Blue, in what is ‘possibly the easiest place in the world to set up an airline’. The profitability situation on the trans-Tasman market is not so rosy, with plentiful capacity keeping pressure on prices e a situation that, from the airlines’ point of view, will have been exacerbated
11 That is, each observation is of the lowest fare available at the time. Of course when that fare was a “Smart Saver” fare tickets could also be purchased in the “Flexi Saver” and “Flexi” fare classes. Thus, when, say, the Flexi fare was observed (usually close to travel date), it means that the lower fare classes or “buckets” had by then been removed from the market. 12 Although flights leaving from locations where it would be inconvenient to load a bar are scheduled to depart at 455pm rather than the 500pm which turns out to be their actual take-off time.
T. Hazledine / Journal of Air Transport Management 17 (2011) 130e135
133
Table 1 Prices observed on Air New Zealand, NovembereDecember 2004. Route
Distance (Kms)
Daily number of flights
SuperSaver fare class
AucklandeDunedin AucklandeNapier AucklandeQueenstown AucklandeWanaka AucklandeWellington ChristchurcheQueenstown ChristchurcheWanaka WellingtoneChristchurch
1062 328 1023 1000 480 347 294 303
12 11 6 1 19 6 1 13
109
129
128
148
90 84 86 70
100 94
139 86 168 155 110 104
169 96 188 185 120 114
80
90
100
since the February 2009 introduction of Emirates’ new A380 superjumbo on its daily AucklandeSydney flight. Nevertheless, Air New Zealand’s relative position seems to be better than the other carriers, in terms of yield per passenger. Table 2 is digested from Hazledine (2008) and shows the coefficients of airline dummy variables in lowest fare regressions (i.e.,using observations like those summarised for Air NZ’s domestic routes on Table 1) controlling for costs, market structure and demand shift factors which are not here shown. The database includes the 2004e05 domestic NZ observations, and also trans-Tasman flights observed in mid 2005 and mid 2006. The coefficients on the airline dummies show the airlines’ prices relative to Air New Zealand’s price, other factors held constant. The three pricing variables are, first, a (weighted) average of all observed lowest fares offered for each flight; second, the lowest fare available eight weeks out, which usually but not always is the lowest fare of all, and, third, the lowest fare available the day before the flight, which for most but not quite all flights is the highest lowest fare observed. The results are quite striking. Overall in these markets Air New Zealand is able to charge a substantial fare premium over its rivals e fares more than ten percent higher than its legacy airline competitor Qantas, and around twenty five percent above either the LCC Pacific Blue or the 5th Freedom carrier Emirates.13 The price differentials are much smaller eight weeks before the flight, when leisure travellers dominate the market. It is in the more lucrative late-purchasing business segment of the market that Air New Zealand is able to pull away from its rivals including Qantas in terms of its yields, being able to sustain price differentials of close to 30% over both Pacific Blue and Emirates. There is little evidence here to support the proposition that LCC competition, granted its overall impact on fare structures, has made legacy carrier pricing practices, and in particular, price discrimination, obsolete. 4. The legacy response II: Air Canada American Airlines’ Saturday Night Stayover restriction, brilliant though it was, came at a cost. It amounts to “spoiling” the product, because it must deter leisure travellers who either do not wish to make a return trip (they want a one-way or perhaps an open-jaw itinerary), or who do not wish to stay away for a Saturday night. Selling cheap one-way tickets, as does Air New Zealand, of course eliminates that element of spoiling of the product, which is no doubt part of its success in increasing sales. The Express Fare system, however, still retains a market-spoiling element from the old legacy pricing regime. Note from Table 1 that there is no overlap between the fare class price points. And recall
13
FlexiSaver fare class
Note that Pacific Blue and Emirates each only competes on a subset of transTasman routes. Also note that the Qantas differential is in real terms even larger than the coefficient implies, because this airline gives FFPs in all its fare classes, and serves hot food and beverages on its domestic NZ flights.
199 106 218 215 150 144 136 120
219 126 248 245 170 164 156 140
239 146 278 275 190 184 186 150
Fully Flexi fare class 279 176 308 315 200
299 206 338
160
180
230 229
399 226 488 375 280 254 242 220
479 246 608 485 330 276 276 260
559 306 708 360 314 290
that the use of the system to implement inter-temporal price discrimination basically involves progressively removing lower fare “buckets” from the market, until, in the last week or so before flight date, only “Flexi” fares may be available. That is, the higher fares come with additional features bundled in. But many leisure travellers, and even some business travellers, may place little or no value on such fripperies as FFP points, status points, and even flexibility to make itinerary changes. Then, if there is any cost in supplying these features, it may be inefficient to bundle them rather than offer a no-frills product as an option. Perhaps in response to these considerations, Air Canada has taken the Air NZ innovations another step e perhaps not a step further, but ultimately in a strikingly different direction. Superficially, Air Canada’s domestic fare offerings look rather like Air New Zealand’s. There are three economy cabin fare classes14 e Tango, Tango Plus, Latitude-with more features and higher prices in the higher classes, and tickets are all one-way, of course. But whereas in the Air NZ’s case one gets the impression that the fare classes really mostly serve as a convenient way to divide up the price points into digestible sets e e.g. three sets of four points rather than a long ladder of twelve points-for Air Canada, the fare classes have become important products in themselves. Insights into the airline’s strategy can be gleaned from various speeches by senior executives preserved on their website. Thus, in May 2006, it was stated that the goal was to have at least five “options” per fare “product” e that is, creating distinct groupings of characteristics, each group developing its own integrity as a recognised product in the market place.15 So, the “products” each have their own slogans on the website, along with suggestive images of the target customers: a young blond woman for Tango (“Our best value”); a laughing black man for Tango Plus (“Get up and go!”); a cool white dude for Latitude (“A perfect fit”); and a handsome mature gentleman for Executive Class (“maximum comfort and freedom”). 16 A basic requirement for any product to have integrity in the mind of consumers is that it be regularly and dependably available, and here we see the first big difference from Air NZ. Whereas it is quite unusual for the latter airline to still offer its lowest Smart Saver fare the day before travel, in the database David Gillen and I have assembled on Air Canada, we find that in about 90% of flights the Tango Plus, if not the Tango, fare is still available the day before, on domestic routes. Now, this fact would not be so interesting if the differences in prices of the fare products were small, but they are not. The premium for Tango Plus over Tango is actually quite low e $35 is
14 In 2006, Air Canada had a fourth economy fare class e Latitude Plus e but this has subsequently been eliminated. 15 http://www.aircanada.com/en/about/investor/documents/sctac-may06.pdf. 16 http://www.aircanada.com/en/about/media/presentations/documents/NBF_ March_28.pdf.
134
T. Hazledine / Journal of Air Transport Management 17 (2011) 130e135
Table 2 Airline dummies in airfare regressions. Sample: 1963 domestic NZ and trans-Tasman flights observed 2004e06
Weighted average of nine lowest price observations
Lowest price observed eight weeks before flight date
Lowest price observed one day before flight date
Dependent variable Qantas PBlue Emirate
Log(Pwavk) 0.117a 0.239a 0.225a
Log(P8k) 0.087a 0.102 0.023
Log(P0k) 0.182a 0.316a 0.314a
a
Significant at the 1% level.
common and the benefits quite substantial: more status miles, more FFP points, access to the Lounges for $30; eligibility to use upgrade certificates to business class, subject to availability. Here the airline is trying to encourage “buy-up”, and in its website presentations it has boasted achieving in 2006 a “48% buy-up”, which presumably means that nearly one third of the passengers purchasing one of the low-fare products paid to get Tango Plus. But the truly striking, even, remarkable feature of Air Canada’s system is that the price differential between Tango Plus and Latitude is huge: around $200, on average, as we will see. True, Air New Zealand’s fare class price differentials are of similar size, but the point is that, as noted above, last minute travellers generally cannot choose between both products e the Smart Saver fares have been taken off the market. So Air Canada, to preserve the integrity of its fare products, is offering seats on the same flight the day before travel at substantially different prices. But why would anyone purchase the more expensive ticket? Well, do they indeed purchase Latitude fares? Again, from the presentations stored on the website, we find the statement: “Tango only accounts for 45% of domestic sales in Quarter 4 [2006]”, which with the 48% buy-up to Tango Plus, implies that around one third of tickets sold were either in business class or were Latitude fares. Now, from the demand side, it is quite hard to match the price differential charged for Latitude over Tango Plus with the additional features obtained. These are: no $50 change fee, complimentary snack, more FFP miles, plus better chances of being able to use an upgrade certificate to the business class cabin. What this probably bears witness to is the agency problem created when business and government travellers don’t have to personally bear the cost of their tickets. Be this as it may, there are clearly large numbers of domestic Air Canada travellers who are willing to pay or to get someone else to pay for various perks and frills which Air Canada can supply and its LCC domestic rival WestJet cannot. Furthermore, looking now at the supply side, it seems that the profit margins on these frills must be substantial. Indeed, the Latitude fare is a product created, almost literally, out of nothing (apart from the cost of flying the customer, which is common to all the fares). It is almost costless to supply these frills, given the non-linearities of operating a full service national network carrier. There are substantial fixed costs and low variable costs in setting up a business class cabin and service, in operating a network of lounges, in setting up and marketing the FFP, in joining the Star Alliance, as well as in operating a full regional network and marketing campaign. For example, the smallest business class cabin operated by Air Canada has nine seats (on its Embraer 175 and 190 aircraft), because of fixed costs. When some of these seats have not been sold to full-fare paying customers, it is virtually costless at the margin to allow high-status or Latitude fare economy customers to upgrade. Indeed, the airline may thereby free up economy seats which can be sold profitably to standby customers. Air Canada’s opaque (in comparison to Air NZ) FFP rewards redemption processes appear to
be managed (as even high-status Aeroplan members know) so as to eliminate opportunity costs of making seats available. And, of course, the non-linearities of the reward schedules effectively lock many fairly frequent Canadian fliers into the Air Canada plan. Access to Maple Leaf Lounges at $25 or $30 probably easily covers the marginal costs of the food and beverages consumed within. So, it seems that Air Canada has come up with a pricing and marketing strategy which allows it to generate quite large revenues from its more-or-less locked in high willingness to pay customers, whilst still covering itself against LCC competition from WestJet by keeping a regular and quite dependable supply of budget fares in the market at all times.17 Table 3 shows how this works out. Here we examined Air Canada’s fares charged in the context of three different market structures: domestic Canada monopoly; domestic Canada duopoly with WestJet, and trans-border routes on which it competes with usually three or four or more US legacy carriers (some serving the routes through code shares, usually with global Alliance partners). To standardise for the very different route lengths, which of course are reflected in fare structures, all the fares are converted to 1400 km journey equivalent, using the coefficient on the log of route distance in an econometric pricing model.18 This is the average distance flown in our sample of routes, which includes 380 Air Canada flight numbers on 46 domestic and 27 trans-border routes, observed weekly from three weeks before to one day before take-off, for flights departing on three consecutive Wednesdays in October 2007. 17 of the domestic routes were then Air Canada monopolies; 29 were shared with WestJet. The first two columns on Table 3 show the (average) Tango Plus and Latitude fares offered 22 days before the flight date; the third and fourth columns show these fares one day before the flight, and columns five and six are the price of the “frills” e i.e., the premium paid for a Latitude ticket over the price of a Tango Plus ticket e 22 and 1 day before the flight. Look first at Air Canada’s behaviour on its nineteen domestic monopoly routes. We see that there was quite a lot of intertemporal price discrimination, with Tango Plus fares moving up 37% over the three weeks, on average.19 However, there is not a lot of change in the frills price, which just slips from $213 to $197. This may not mean a lot, because most business travellers don’t appear in the market until a week or ten days before the flight, so that the Latitude fare 22 days out is probably not heavily sold. Now look at the duopoly routes. It is strikingly clear that competition makes a difference to Air Canada’s pricing e its Tango Plus fares are just about two thirds as high as on the monopoly routes. But is it “low-cost” competition that matters? Possibly not. The frills markup is a bit lower in dollars per km; a bit higher as a percentage, in the duopoly markets. That is, the pricing game that Air Canada is playing with its “fare products” does not seem to be
17 Although the Tango or Tango Plus fare was not available day before for 10% of Air Canada’s flights in our database, it was almost invariably possible to purchase a cheap ticket on some e usually mid morning or afternoon or red-eye e flight on the same route the day before, and it seems that that fare was typically the same as the lowest available WestJet fare on the route. Thus, the proud quoting by Air Canada executives of their rival WestJet’s CEO Clive Beddoe’s statement: “Air Canada matches us dollar for dollar on every single fare” is not exactly false, though it could be misleading [http://www.aircanada.com/en/about/media/presentations/ documents/NBF_March_28.pdf]. 18 The coefficient on route distance was 0.75 in a model with price per kilometre dependent. That is, other things equal, the actual fare (ticket price) increases by 25% for each doubling of route distance. 19 Although it is not demonstrated in the average figures of Table 3, it is quite common for the day 1 Tango Plus fare to exceed the Day 22 Latitude fare for a particular flight. This is another difference with Air New Zealand’s Express system, where the price points always move up monotonically through the fare classes.
T. Hazledine / Journal of Air Transport Management 17 (2011) 130e135
135
Table 3 Air Canada’s Strategic Product Enhancement. Average fares and prices of frills, adjusted to 1400 km journey ($)
Canada monopoly routes Canada duopoly routes Trans-border routes All routes
Tango þ fare, 22 days before take-off
Latitude fare, 22 days before take-off
Tango þ fare, 1 day before take-off
Latitude fare, 1 day before take-off
Price of frills 22 days before take-off
Price of frills 1 day before take-off
473 302 309 336
686 480 517 529
648 430 767 564
846 619 858 722
213 178 209 193
197 189 91 158
Note: Data from Gillen and Hazledine (forthcoming), prices observed on Air Canada website for 46 domestic Canadian and 27 trans-border routes; 380 flight numbers; each flight observed four time from 22 days to 1 day before take-off, on three Wednesdays in October, 2007.
much affected by the presence or otherwise of its LCC domestic rival, WestJet. The contrast with the trans-border routes is quite dramatic. With the exception of a couple of Florida destinations (Orlando and Tampa) to which WestJet flies, Air Canada is now getting its competition from a bevy of US-based legacy carriers. These are airlines which have not changed a great deal in their marketing practices. They still offer discounts on return tickets and on Saturday Night Stayovers20. Their one-way ticket offerings are quite rudimentary e you can buy a refundable or non-refundable ticket. So, with the exception of the Florida routes, Air Canada transborder does not offer Tango fares. And, although the Tango Plus/ Latitude difference 22 days out is in line with the domestic differences, when we get to the business end of the market close to the flight date, the differential is sharply compressed. Basically, Air Canada matches the refundable/non-refundable fare differences charged by its US rivals, these being on some routes as low as $20. The fact that Air Canada can do this, and presumably not lose money, is evidence on just how cheap the Latitude frills must be to supply. 5. Conclusion These case studies document two legacy airlines’ innovative and spirited responses to the low-cost carrier challenge, and demonstrate, at the very least, that even well-run LCCs are not a force majeure, sweeping aside all resistance. The simplicity which is the strength of the low-cost business model is also a potential source of weakness, that can be exploited by a legacy carrier determined to extract full value from its unique fixed assets. There may, however, be some doubts about the applicability of the Air Canada and Air New Zealand strategies to the situation of legacy carriers in larger markets such as the US and Europe. These two airlines have the advantages of national carriers, which include a myriad of formal and informal business linkages within their home market; complete regional coverage of this, and the capacity to operate on long-haul routes where LCCs cannot follow. Thus, it is quite striking that Air Canada is apparently unable to extract the same markup on its premium fare class on trans-border routes, where its competition is other legacy carriers, as it earns domestically in competition only with WestJet. The practices of the US legacy
20 Puller et al. (2009) find that round-trip and SNS tickets come with 11% and 13% discounts, in the US.
carriers, however, which still impose return-journey restrictions on their low-fare tickets, are coming to seem rather antediluvian, and it is just possible that the examples of two “foreign” carriers’ quite successful responses to LCC competition might give them the nerve to try something innovative themselves.
Acknowledgements Without implication, the author thanks two referees and participants in the 2009 GARS Workshop, the 12th Hamburg Aviation Conference and the Sullberg Society 3rd Meeting for comments and insights, especially Nicole Adler, Achim Czerny, Hans-Martin Niemeier, Mike Tretheway, Martin Dresner, David Gillen, Ken Button, Anming Zhang and David Starkie. Some of the material used in this paper is drawn from research in progress joint with David Gillen.
References Berry, S., Jia, P., 2009. Tracing the Woes: An Empirical Analysis of the Airline Industry. Annual Conference. American Economics Association, San Francisco. Button, K.J., 2004. Does the theory of the ‘core’ explain why airlines fail to cover their long-run costs of capital? Journal of Air Transport Management 9, 5e14. Dana, J.D., Orlov, E., 2009. Internet Penetration and Capacity Utilization in the US Airline Industry. Annual Conference. American Economics Association, San Francisco. January. Dresner, M., Lin, J.-S., Windle, R., 1996. The impact of low-cost carriers on airport and route competition. Journal of Transport Economics and Policy 40, 473e571. Goolsbee, A., Syverson, C., 2008. How do incumbents respond to the threat of entry? Evidence from the major airlines. Quarterly Journal of Economics 123, 1611e1633. November. Gillen, D., Gados, A., 2008. Airlines Within Airlines: Assessing The Vulnerabilities of Mixing Business Models. Research in Transportation Economics 24, 25e35. Gillen, D., Hazledine, T., forthcoming. The New Pricing in Canadian Air Travel Markets: Implications for Competition and Antitrust. In: James Peodes (Ed.), Advances in Airline Economics, Vol. 3, Emerald Books. Working Paper. Centre for Transportation Studies, University of British Columbia. Hazledine, T., 2008. Modern Airfare Price Discrimination: Theory and Evidence. University of Auckland. Hazledine, T., Pricing., 2010. competition and policy in Australasian air travel markets. Journal of Transport Economics and Policy 44, 37e58. Puller, S.L., Sengupta, A., Wiggins, S.N., 2009. Testing Theories of Scarcity Pricing and Price Dispersion in the Airline Industry. Annual Conference. American Economics Association, San Francisco (January). Tretheway, M., 2004. Distortions of airline revenues: why the network airline business model is broken. Journal of Air Transport Management 10, 3e14.