WEB APPENDIX D CAPACITY PLANNING AND PRICING AGAINST A LOW-COST COMPETITOR: A CASE STUDY OF PIEDMONT AIRLINES AND PEOPLE EXPRESS

Transcription

1 WEB APPENDX D CAPACTY PLANNNG AND PRCNG AGANST A LOW-COST COMPETTOR: A CASE STUDY OF PEDMONT ARLNES AND PEOPLE EXPRESS ARLNE ENTRY STRATEGY During early 1981 People Express (PX) became one of the first new entries in the deregulated interstate airline industry. PX s entry strategy was offer a uniform low-price, no frills, high-frequency regionwide service 13 peripheral mid-atlantic cities using a hub and spoke system out of Newark, New Jersey. By unbundling all services, adopting quick turnaround times, working longer crew shifts, and converting all first-class and galley space in additional coach-class seats, PX achieved a 31 percent reduction relative the industry average in direct fixed costs per flight (e.g., crew costs) and a 25 percent reduction in variable costs per seat (e.g., cabin service). Having secured the lowest operating cost structure in the industry, PX set out attract cusmers who saw air travel as a commodity and would regularly fly rather than drive. The protypical target cusmer was a manufacturer s trade representative who often needs travel on short notice, but is seldom on the company expense account. n essence, People Express created a new segment of the market not previously served by much more expensive and infrequent Mohawk and Allegheny flights (the predecessors of US Airways). As a result, inverse intensity rationing of the cheap capacity ensued; that is, the new low-willingness--pay cusmers attracted in the market by PX s discounting quickly secured all of PX s capacity, leaving almost none available other air travelers. As a result, PX failed take regular cusmers away from the higher-priced incumbents. Figure WD.1 displays the strategy game this entry presented the mid-atlantic regional airlines. The incumbents had decide whether match PX s deeply discounted fares or accommodate PX by maintaining high fares. PX had decide whether enter with a large-capacity 120-seat Boeing 737 or a small-capacity 30-seat dehavilland 128. LARGE-SCALE ACCOMMODATON n Entry Deterrence in Chapter 13 of the textbook, the new entrant had decide first. n the case here, prediction of the equilibrium strategy was easy because PX had a dominant strategy. No matter what the incumbents pricing response in the mid- Atlantic markets, PX had higher payoffs from a large-capacity entry. PX s payoffs of $3,877 and $2,068 per flight with frequently scheduled larger capacity aircraft dominated $969 per flight with frequently scheduled small-capacity aircraft. Examining the payoffs resulting from large-capacity entry, the incumbents preferred accommodate with $8,225 operating profit per flight rather than match and earn only $5,368 per flight. Even though by matching PX s discount, incumbents could pick up WD-1

2 WD-2 WEB APPENDX D Capacity Planning and Pricing Against a Low-Cost Competir Figure WD.1 Large-Scale Entry Accommodation with nverse ntensity Rationing Excess Capacity PX Precommitment Small-Scale Entry Large-Scale Entry Match Low Price Match Low Price ($4,236, $969) ($4,765, $969) ($8,225, $3,877)* ($5,368, $2,068) Note: Profit-per-flight payoffs are listed as (Mid-Atlantic incumbent) and PX (People Express entrant). extra seats in overflow demand from People s cusmers, the loss of margin was sufficient induce the incumbents prefer accommodate large-scale entry. {Large Capacity, } therefore proved be the subgame perfect equilibrium. Using leaseback purchases leverage its modest start-up capital in 17 largecapacity planes, People Express intensively developed a dozen peripheral mid- Atlantic routes and thereby managed avoid retaliation from either the regional or major carriers. n the Southeast regional markets, the situation was entirely different. Piedmont Airlines was the second-fastest-growing airline in the post-deregulation era. By serving small- medium-sized cities ignored by the major carriers and by connecting through hubs with little or no competition, Piedmont retained a record 95 percent of its passengers on connecting flights. Piedmont knew that its reputation with business travelers was high and rising, but also knew that measuring tactical success against a new entrant would require hard data. Piedmont therefore proceeded count and categorize every passenger on every PX flight in Piedmont cities. From these competir surveillance data, Piedmont determined that several travel segments sorted randomly the low-price supplier when substantial price differentials were present, but loyalty the incumbent prevailed when prices were identical. This brand preference for the incumbent was almost universal at $79 prices, but remained strong even with the new lower-willingness--pay cusmers attracted in the market by People s discounting $49, $39, $29, and even $19. LARGE-SCALE ENTRY DETERRENCE Figure WD.2 displays the entry deterrence and accommodation game with random rationing that presented itself Piedmont and People Express in early 1985; the boxed data list the common information. On a typical 400-mile route, Piedmont had decide whether match PX s $49 one-way fares or accommodate the new entrant by maintaining its own $79 one-way fares. Again, PX had decide whether enter with a large capacity 120-seat Boeing 737 or a small capacity 30-seat dehavilland 128. And again, PX had decide first; the incumbent s (i.e., Piedmont) pricing decision controlled the endgame.

3 WEB APPENDX D Capacity Planning and Pricing Against a Low-Cost Competir WD-3 Figure WD.2 Matching Price Response with Random Rationing Common nformation Starting price $79 New price $49 Demand at $79 Demand at $ ncumbent capacity 300 Variable costs (P) $16.69 Variable costs (PX) $13.63 Small Capacity P (30 seats) PX Scale of Entry Decision Large Capacity P (120 seats) (P = $79) Match (P = $49) (P = $79) Match (P = $49) Total Demand 196 Sales (P) Sales (PX) new in 64 market new in 64 market 0 64 ($6,979, $1,061)* ($6,330 $5,365, Profit (P, PX) $0 $1,061) ($3,178, $4,244) ($6,330 $0 $4,265, $2,264) Notes: Payoffs are listed (ncumbent, Potential Entrant). P refers the incumbent airline (Piedmont) and PX refers the potential entrant and actual entrant airline (People Express). The sales ranges and payoff ranges are described in the text. Analysis of the subgame perfect equilibrium strategy is straightforward. First, let s examine the payoffs resulting from small capacity entry in the p half of the diagram. The incumbent (P) prefers accommodate with $6,979 operating profit per flight rather than match and earn only $5,365 per flight. Even if by matching PX s discount, the incumbent picks up 30 extra seats from People s cusmers, the loss of margin from ($79 $16.69) ($49 $16.69) is sufficient reduce the operating profit from $6,979 $6,330. Consequently, the second row of endgame outcomes must be eliminated by PX from further consideration; Piedmont would never select it. Turning then the large capacity payoffs, the situation is a bit different. With its 120-seat capacity, a 737 PX can satisfy all 64 new demanders attracted in the market at price points between $79 and $49 and, in addition, can offer 56 seats of remaining capacity the incumbent s price-sensitive cusmers. n PX s mid-atlantic markets, this had not proven a problem for incumbents because inverse intensity rationing of PX s cheap capacity described the cusmer sorting pattern. n contrast, in Piedmont s Southeastern markets, PX s capacity was randomly rationed. With accommodative differential prices of $79 and $49, every cusmer willing pay at least $49 would then have an equal probability of securing service from the lowprice entrant. For large-scale entry, this meant that 120 seats would be rationed among 196 cusmers so each of the cusmers willing pay the incumbent s

4 WD-4 WEB APPENDX D Capacity Planning and Pricing Against a Low-Cost Competir $79 accommodating price faces a (1 ( ) 196 ) 0.61 chance of being served and a 0.39 chance of being denied service at the low-price entrant. Therefore, the expected demand at Piedmont from accommodating large-scale entry was seats, which implied an expected operating profit of only $3,178. Responding the large-scale entry by matching prices offered the potential for much greater Piedmont profit (i.e., $6,330). Again, ascertain the payoffs in the face of large-scale entry, the incumbent employed extensive competir surveillance and tracking of target cusmers, such as the regional headquarters personnel flying regularly back and forth corporate headquarters in New York. These data allowed Piedmont conclude that at worst, with matching $49 prices, People Express would attract all 64 new cusmers who had entered the market in response discounting, but none of Piedmont s regular cusmers. At worst, then, Piedmont would receive ($49 $16.69) $4,265 from matching PX s discount and only $3,178 from accommodating prices. Consequently, PX should have eliminated from further consideration its otherwise attractive $4,244 per flight outcome associated with Piedmont s accommodation of PX s large-scale entry. Predictably, Piedmont would never go for it. Then, from PX s perspective, the profit per flight from large capacity entry, using PX s lower incremental variable cost of $13.63, was at best ($49 $13.63) 64 $2,264 and could go as low as zero. n contrast, the profit per flight from small capacity entry was a nearly certain $1,061. PX should have realized that the subgame perfect equilibrium was {Small Capacity, }. nstead, what actually happened was that People Express entered with the same large capacity employed in its peripheral mid-atlantic routes. Piedmont was surprised, but of course matched prices, and a price war quickly ensued. At one point, the fares got down $19, and residents of the Carolinas found themselves invited tavern lunches at Tavern on the Green in New York. Eventually PX s operating profits fell so low it was forced withdraw. People Express appears have failed recognize that with random rationing at differential prices and loyalty the incumbent at matching prices, only small-scale entry could have induced accommodation and the sustainable $1,061-per-flight payoffs that were available in this market. This basic insight of several entry deterrence and accommodation games is sometimes humorously referred as judo economics. The subgame perfect strategic equilibrium in which mid-atlantic incumbents were induced accommodate large-scale entry presumed inverse intensity rationing, such as PX had experienced in its original markets. When the cusmer sorting pattern changed, the subgame perfect strategic equilibrium changed. People Express chose the right strategy, but for the wrong game and ended up being deterred. The fundamental and generalizable managerial insight here is that differences in cusmer sorting patterns alter the best-reply responses of incumbents facing entry. Exercises 1. n the entry deterrence and accommodation game between Piedmont and People Express, explain why Piedmont s expected demand following small-scale entry in Figure WD.2 is What cusmer sorting rules likely apply at the local airport where you reside? Do they differ in your estimation across destinations and target cusmer segments? Why or why not?

5 WEB APPENDX D Capacity Planning and Pricing Against a Low-Cost Competir WD-5 3. Why is knowing just one dominant strategy in the People Express entry game sufficient predict rival behavior and identify the strategic equilibrium? 4. Today United Airlines and Delta Airlines face competition from discounters JetBlue and AirTran. One important difference, however, is that AirTran and Delta have virtually identical costs of 8.6 and 8.8 cents per available seat mile, whereas JetBlue, like People Express, has costs much lower than the incumbent. t costs United $23,690 operate an Airbus 320 cross-country from Dulles in Washingn, D.C., Oakland, California, with exactly the same crew and equipment that JetBlue operates for only $14, Booz Allen estimates that 39 percent of this huge cost differential is attributable processes required by United s hub-and-spoke system as opposed JetBlue s point--point operations with a single type of aircraft. Also, JetBlue pilot costs are 25.5 percent of tal costs (lowest in the industry), whereas United s are 49.7 percent (highest in the industry). Devise an entry deterrence or accommodation strategy for Delta and for United. 1 Costly Race in the Sky, Wall Street Journal (September 9, 2002), p. B1.